CreditAccess Grameen Limited (CREDITACC) Earnings Call Transcript & Summary

Ladies and gentlemen, good day, and welcome to CreditAccess Grameen Q4 FY '24 Earnings Conference Call. [Operator Instructions] Please note that this conference is being recorded. I now hand the conference over to Mr. Nidhesh Jain. Thank you, and over to you, sir.

Thank you, Sagar. Welcome to the Quarter 4 FY '24 Earnings Conference Call of CreditAccess Grameen Limited. To discuss the financial performance of CreditAccess Grameen and to address your queries, we have with us today Mr. Udaya Kumar Hebbar, MD of CreditAccess Grameen; Mr. Ganesh Narayanan, CEO of CreditAccess Grameen; Mr. Balakrishna Kamath, CFO; and Mr. Nilesh Dalvi, SVP and Head, Investor Relations. I would now like to hand over the call to Mr. Udaya Kumar Hebbar for his opening comments. Over to you, sir.

Thank you, Nidhesh. Good evening, everyone, and we welcome you to join the conference call to discuss our fourth quarter and full year FY '24 business performance. CA Grameen has stood as a preferred financial partner supporting the entrepreneurship journey for millions of women over past 25 years. We take pride in being the first stand-alone microfinance entity to surpass the significant milestone of INR 25,000 crores AUM mark in our Silver Jubilee year of operations. We have been once again certified as a Great Place to Work for the fifth consecutive year given our purpose-oriented approach. We ended financial year '24 on a solid footing, having reached an AUM growth of 25% Y-o-Y and 14.2% Q-on-Q to INR 26,714 crores, supported by a healthy increase in customer base. The customer base grew 15.3% Y-o-Y and 4.8% Q-o-Q to 49.2 lakh at the end of financial year '24. Our customer addition momentum continued to remain strong as we added over 13.5 lakh new customers during FY '24 and 4.2 lakh new customers in Q4 FY '24 with an average of 1.1 lakh addition every month. The expansion in outside core markets have been playing out well as 34% of our yearly customer addition are outside of top districts. We added 86 branches in the last quarter and 194 branches during this year, taking our total infrastructure strength to 1,967 branches spread across 383 districts at the end of March '24. We also merged 13 branches to enhance our operational efficiency. Once again, disbursement was the strongest in Q4 FY '24 at INR 8,053 crores, an increase of 12.3% Y-o-Y given robust customer additions and renewals gained from the retained borrowers. Our progress in retail finance portfolio continued to thrive with an AUM of INR 708 crores at the end of March '24, an increase of 320% Y-o-Y. This book, being 2.7% of our overall portfolio, is performing extremely well with PAR 30 of around 0.07% across new products, like individual business loans, mortgage-backed loans, 2-wheeler loans, et cetera. In line with our evolving customer philosophy, we are supporting our value-added customers to retail finance product, resulting in an industry-leading customer retention rate of 88%, a proof of strong customer affiliation with our brand. Net interest income in Q4 FY '24 grew by 33.7% Y-o-Y to INR 922 crores. We continue to remain one of the lowest-cost lenders with the portfolio yield of 21%, a stable cost of borrowing at 9.8% and operating with 11.2% interest spread. Our diversification in liability franchise both source and geography-wise, the highest standalone credit rating, adequate ESG disclosures and ratings, and healthy capital adequacy contribute to our unique positioning. We believe that our cost of borrowing for FY '25 should remain stable as we continue to maintain a robust ALM position. And our interest spread was 11.2% in Q4 FY '24. NIM stood at 13.1%. Cost-to-income ratio was healthy at 30.1%, which is 10 bps lower than Y-o-Y. We created an additional provision of INR 26 crores towards long-term incentives and onetime special bonus to our employees while celebrating our Silver Jubilee year. PPOP grew by 35.8% Y-o-Y to INR 683 crores. The credit costs in Q4 FY stood at INR 153 crores, partially offset by INR 13 crores of bad debt recovery. The collection efficiency, excluding arrears, stood healthy at 98.3%, GNPA at 60 DPD stood at 1.18% and PAR 90+ at 0.94%, and net NPA at 0.35%. ECL provisioning stood at 1.95%, higher than our GNPA. Our PAT grew by 33.9% Y-o-Y and 12.4% Q-o-Q to INR 397 crores during Q4 FY '24, resulting in ROA of 5.7% and ROE of 24.9%. For the financial year '24, net income grew by 45.9% Y-o-Y to INR 3,260 crores, while PPOP grew by 58.7% Y-o-Y to INR 2,391 crores. Gross credit cost stood at 2.06% in FY '24. As explained on Slide 7 of the investor presentation, the credit cost on account of delinquent loan was 1.71%, which included 1.52% of actual credit loss in the form of write-offs and 0.19% provision in Stage II and Stage III provisioning. The remaining difference of 0.35% was on account of increase in standard asset provisioning on Stage I assets basis our revised ECL policy introduced in FY '23. Our Stage I provisioning increased from 0.82% in March '23 to 0.92% as on March '24. The net credit cost after incorporating INR 48 crores bad debt recovery stood at 1.8% for FY '24, in line with our guided range. The company continues to hold over and above 100 bps higher provisioning compared to IRAC prudential norms. Our PAT in FY '24, grew by 75% Y-o-Y to INR 1,446 crores, leading to an ROA of 5.6% and ROE of 24.9%, in line with our annual performance guidance. Capital adequacy remained comfortable at 23.1%. It is important to note that our cross-cycle ROA and ROE over the past 8 financial years, from FY '17 to FY '24, despite facing 3 external shocks, stood at 3.6% and 15%, respectively. Our distinctive methodology for conducting the business in the form of classical JL model, diverse product suite, flexibility in repayment and continuous district-based expansion approach, supported with strong corporate governance structure, ensures stability and enhanced resiliency. Commemorating our 25 years of operations, the Board of Directors have recommended the onetime final dividend of INR 10 per share, the dividend payment of INR 159 crores translating to a dividend payout of 11%, subject to shareholders approval at the ensuing Annual General Meeting of the company. Assuming a stable operating environment, we look forward to achieving loan portfolio growth of 23% to 24% in FY '25. On the back of our competitive lending rate and strong control on the cost of borrowing, we foresee NIM in the range of 12.8% to 12.9%. The cost-to-income ratio should be range bound at 30% to 31%. We are anticipating credit cost of 2.2% to 2.4% with the actual write-off being within 2%. Overall, we aim to achieve ROA of 5.4% to 5.5% and ROE of 23% to 23.5% in FY '25. Looking at the medium term from '25 to '28, we anticipate loan portfolio growth of 20% to 25%, NIM in the range of 12.7% to 12.9% and cost-to-income ratio between 30% to 31%, helping keep our leadership position intact. We expect a similar credit cost of 2.2% to 2.4% while delivering ROA of 5.4% to 5.5% and ROE of 22% to 23.5%. With this, I would like to open the forum for question and answers. Thank you for your support and your listening.

[Operator Instructions] The first question is from the line of Renish from ICICI.

Congrats on a good set of numbers. Sir, just 2, 3 questions from my side. One on the clarification side. So on a sequential basis, our emergency loan proportion has spiked up significantly to INR 1,000-plus crores. So can you please throw some light on what sort of emergency loans these are?

I think there are some data, we will look. It's less than 0.8% of our book, actually. I think there's some data that we'll report again, Renish.

Renish, the emergency loans is INR 5 crores.

INR 5 crores only, total portfolio.

We rectified it in the deck.

Got it. Got it. I think -- so there is a mismatch between home improvement and emergency, I guess.

There seems to some error, we will rectify that.

Okay. Okay. No, no. No worries. Sir, secondly, on the margin side, of course, this year has been fantastic for us, but now going ahead, when we say that we'll be maintaining this 12.7% to 12.8% NIM and the way sort of regulator is commenting on the yields, do we foresee any risk to sustain these margins going ahead?

So now, if you look at our financials, our interest spread is only 11.2% currently. It's quite reasonable when compared to the regulatory direction also. As well as we are the lowest price to customer. And the NIM is coming from the efficiency of managing capital and managing operations. We don't see a reason where this is seen negatively. I think this is -- this will be seen positive.

Got it. Got it. And sir, just last question, maybe a bit lengthy one on the credit cost side. So we were guiding for 2.1% of credit cost for '24. Of course, we ended very close to that. But then when we look at the increase in guidance from 1.6% to 1.8% to 2.1%, and now 2.2% to 2.5%. So sir, what has changed fundamentally which is sort of leading to this increasing guidance for credit cost maybe every 6 months?

Yes. There are two things are happening. One is actually grow more in the noncore market. Our provisioning is higher because we provide them at a high-risk rate. For standard provisioning -- the impact of increased credit cost is basically almost 35 bps coming from the higher provisioning, what we're creating for a standard asset, as I explained earlier, also numbers available in our deck. So our actual credit cost is actually 1.7%. So as we grow more in the loan market, there will be some impact on the credit cost. And two, we also observing a slight stable credit cost, it's slightly higher than what it was compared to Q1 and Q2, so which we also explained in the last quarter. So the stable credit cost will remain around this. And that's why the actual write-off, I said maybe less than 2%, but credit costs may hover between 2.2% to 2.4%, considering early recognition of risks and higher provision on standard assets.

Okay. So basically, you're saying, let's say, incrementally between '25 to '28, our growth will be coming largely from the non-southern market or in a sense non-core markets. And given our ECL model is, the Stage I provisioning will be always higher vis-a-vis the historical trend? Is that the current understanding, sir?

Correct. Correct. If you see our Stage I provisioning, which was less 80 bps, moved to more than 90 bps right now. So the difference between last one year, 45 bps of increase in Stage I only because of this change. So basically, there's a cushion being built by recognizing early and then having sufficient cushion from the -- against the risks, potential risks.

Got it. Got it. And again, a follow-up on that is, let's say, so basically, you are trying to highlight that the credit behavior of the customer has changed post-COVID, and which might lead to the little higher steady-state credit cost than what we have seen pre-COVID?

Correct. It's a combination of 2 things, as we said. There will be a little higher steady credit cost and a little higher provisioning itself. The combination of that will show up as credit cost.

So does this reflect in the center attendance as well, sir? I mean, what are the parameters to track this? I mean, a change in customer behavior?

So there's not too much change in customer behavior, that the core market to non-core market I was telling actually. So non-core markets have a little different tendency. Therefore, it will potentially go up a bit as a steady state.

Okay. And what explains the sequential increase in the PAR portfolios across the buckets, sir?

There is only 20 bps increase in our portfolio if you observe. Almost 10 bps is coming from the flood situation still continuing because it is staying on and off, most of them are paying still in the PAR bucket. So other 10% of it just a simple movement here and there.

Okay. Okay. Got it. Got it. But nothing -- let's say, any stress building up in the sector you see? Right?

At least we are not seeing that in any of our markets. There will be a small variation between different states, but not significant changes.

Our next question is from the line of Rajiv Mehta from Yes Securities.

Congrats on strong performance. Just to harp on credit costs, just two clarifications on the guidance that you've given of 2.2% and 2.4% for FY '25. Is it gross or is it net of bad debt recoveries? That's first. And second, does it assume...

Gross, gross.

It is gross, okay. Does it assume any further increase in provisioning rate on Stage I and Stage II? I know we are currently at 92 basis points and 55% (sic) [ 55 bps ] for Stage I and Stage II. Do we intend to take this further up?

That could be because now actually we moved to a district-level recognizing the ECL actually. If a district looks higher, a higher risk or a higher delinquency bucket, high risk district, then there will be higher what you call standard provision also. So there will be small movement -- there could be small movement if there's an increase in the trend, but that will get countered with the pricing, right? Then we're also moving to a district level pricing also, so we'll actually manage between both pricing as well as risk. So there could be a small movement here and there, but overall, we should not see too much of variations.

And just a follow-up, two questions on this borrower attrition for the year was 578,000 borrowers. Can you point out was it concentrated in certain markets that will let go of borrowers because of leverage or because of competition? That's one. And second is there is a meaningful sequential increase in GLP per borrower in higher-vintage buckets of 3 to 6 years and more than 6 years. So is it driven by cycle migration or something else?

One is our attrition is, actually, it's not changed. It's 88% earlier also, it is now also. There is no change in the trend actually. So within that, which is what we hope to maintain between 85% to 90% of the retention. We are within that range. So there's no significant variation in our bucket. There is one variation we saw, that the Madura book where we lend or renewed, some of them probably a little higher in terms of attrition because they all moved from monthly to weekly, some of them may not accept weekly, they want to go back to monthly also. So we saw a little higher there. But otherwise, there is no significant change in anywhere. But overall, it's very insignificant. Sequential increase -- sorry, the sequential increase. Average increase is basically I would say, and this only at a more than 3 years to 6 years bucket only where we have a very strong, loyal customer base only where we have seen sufficient vintage. And the risk also lesser. If you see, we provide lesser -- we've always seen lesser risk in that bucket. So it's a normal demand and supply. So this kind of increase always help, but whereas if you see below 3 years, it's consistently same trend.

And it's anyway less than 10%.

Yes, it's overall 8%. Correct. So if you see, annual growth of GLP used to grow between 9% to 10%, within that range only now also, it's about 8% overall increase.

The next question is from the line of Abhijit Tibrewal from Motilal Oswal.

Congratulations on a good quarter. Just one question. Just trying to understand whether in the presentation also, we say that some bit of the PAR accretion that we are seeing is happening because of stronger growth outside our core states. And you also suggested that, as part of increase in our credit cost guidance, some bit of it in terms of higher write-offs is going to come from our growth in the non-core states. So if you could just provide some more picture with regards to particular states where we are seeing that the PAR accretion is higher. Or is the, I would say, credit behavior of your non-core state customers very different from your core state customers?

Correction, it's not a PAR increase. We said credit cost increased because we provide higher considering that new markets, we are positioned at a little high risk and provide higher even for standard as well as Stage I, Stage II also. Therefore, the credit cost is going up a little bit, not necessarily the PAR is going up. You see, PAR has gone up only by 20 bps, between last 12 months. 10 bps coming from the flood impact, only 10 bps in this because of the overall change. It's not very significant in case.

I was trying to talk more about credit costs.

Credit cost, more of a provisioning itself is higher in the non-core market. That is why.

Got it. And any particular reason? I mean, typically sir, I mean, most of the lenders talk about the [ PVNGD ] experiences.

Old markets, core market where we know the behavior well over a period of time. So our -- we are able to assess the risk properly and see at least more consistently. Whereas new markets, we would take a little more time to assess carefully and see whether it's low risk or high risk or medium risk. So we provide, based on either medium risk or high risk, not the low risk, for the new market at this point of time. Therefore, the credit cost or provision, what we make, is a little higher in the new markets.

Got it. If I understood you right, I mean, this is a conscious choice that we have made, of making higher provision on the newer markets, right? Rather than our ECL models asking us to do that.

Okay. So for example, our Maharashtra actually, we treat it as a high risk. We actually provide higher Maharashtra, Madhya Pradesh, Chhattisgarh, Rajasthan, Punjab, all of them we provide as a high-risk states. Whereas states which are in UP, Bihar and all, we provide as a medium risk. So Karnataka, we provide as low risk state. Tamil Nadu we provide as medium risk state. So we have a differentiate. That from there, we are moving to district now, so now we have about 29% of our districts are in high risk. So moving forward, we actually granular -- we look at the granular as a district level, which the high risk and the low risk. So there will be some change of patterns in terms of providing, but still there won't be too much variation in this process.

Okay, sir. And sir, then to conclude that, I mean, just like we are doing risk-trading based on districts, the pricing is also, I would say, risk-based in each district?

Absolutely. We are moving towards that direction.

The next question is from the line of Shreepal Doshi from Equirus.

Congrats on a good set of numbers, sir. So my question was firstly on this increase in marginal cost of -- sorry, decline in the marginal cost of borrowing. So what is it that is helping us on the cost of funds, sir?

Okay. And I will ask Kamath to respond this. So only thing is majority of our borrowing in Q4 was done domestic only. That is one of the reasons. Maybe Kamath would give a...

Normally in the fourth quarter of the year, we borrow from all the domestic banks and we want to utilize all the sanctions. So we have not drawn any NCD or we have not drawn any foreign borrowings. So that's why the average marginal cost of borrowing has come down in the fourth quarter.

Okay. So for the whole year, it should be broadly in the range of 9.8% sort of a percentage for FY '25 as well if the rates were to go down?

Yes.

The rate may not go down from here because we still continue to diversify our business, diversify our borrowing to NCDs, external borrowing, everything. So it will remain stable at this level unless major change in the lending cost from RBI side or from Fed side.

Yes, you may see some divergence between the quarters based on how we draw the funds. But on the average, we feel it should be largely stable, till the Fed and RBI signal a downward movement on the rate.

And it will be passing on to us.

Yes. That will take some more time because it's all MCLR-based. Our loans are not on the repo. So until we get the benefit of the lower MCLR, the rates could be largely stable.

Got it. Got it. Sir, one question was pertaining to pricing and linking it with your earlier commentary on states like, say, high-risk and medium-risk. So what is our pricing in say high-risk states and medium-risk states versus Karnataka? That was question number one. And question number two was pertaining to asset quality again. So on that front, if you look at the -- I mean, the AUM per loan officer has been increasing. So is that a reason or could that be a reason which would be -- which could explain this deterioration in the high-risk state? And also if you could highlight the attendance in the high-risk states and the attrition in these high-risk states.

Before that. So firstly about pricing, you asked. There are two types of pricing we do based on risk actually. One is risk-based pricing based on what district risk status, high or low. That is one. Second is based on the vintage of the client. With the high-vintage client, the pricing is low, whereas low-vintage client will price higher. Difference between low and high is almost about 2%. Yes, 20.5%, 1.5%. 20.5% is the lowest and the highest is 22.25%. The 1.75% variation between these two. That is point number one. About the...

GLP per branch...

GLP per branch, Kamath?

So with respect to the increase in GLP per branch or loan officer the right metric that we watch is borrowers per a loan officer, right? So if you see the borrowers per loan officer, it still ranges around 375. Typically, a mature market for us can handle around 550 customers on an average, right? As long as the borrowers are in control, increase in GLP is only a factor of vintage and along within a particular geography. So that is not the reason for increase...

And then particularly in the newer market, our customer -- I mean, handling is lesser, not the high. Higher will be handled with the core markets. There's no correlation between these two.

Okay. And then the last part on, say, attrition at employee level in the high-risk. And also if you could give some color on the attendance in these states at the center meeting.

So overall attrition is quite under control. In fact, for us, even in newer geographies, attrition is far better, right? So it is lower than your mature markets. And with respect to center attendance, we've not seen much of a change. But if say 80% is something that we normally prescribe, probably we will be anywhere between 65% to 70%, depending on geographies. And specifically in lower -- sorry, high-risk geographies where the penetration is lower, et cetera, we would demand a slightly larger attendance. And that's something that varies, again, from state to state.

Not only attendance, we'll have a higher control and audit also in the newer markets than the high risk states, so which will help us. When we designed the district-wise high-low rates, we also can trigger the different controls in the district also. So we have that ability also. Right now, also new markets will have higher audit controls, higher risk controls, so which is always a part of our control in attendance.

Got it, sir. All right. Sir just last question, if I can squeeze in, which would be, which states would be the key states for us to deliver -- to help us deliver this 23%, 24% growth in FY '25?

See, our core market will deliver between 10% to 15%. So basically, the other market, which is more of a Bihar, UP, West Bengal, Orissa; and the new market, which is Telangana, Andhra, should deliver higher than 25%. So that will be on an average 23%, 24% growth.

The next question is from the line of Balkrushna Vaghasia from Axanoun Investment Management.

Sir, my first question is on a global level regarding the sector. So as the results are coming for fourth quarter for all the micro finance companies, so everybody is projecting, some say 20% growth for next 3, 4 years, 25%, 30%, even higher than that. So what are the signs that may or may not be present right now, but in future, we should look for when there is overheating in the sector or there is a high kind of optimism and which can lead to some sort of correction in the sector?

Yes. Our view is, in fact, we declared last year that we can grow at least between 20%, 25% for next -- by FY '28, we said what we can reach based on the potential in the market, based on the retention, what we can -- our customer retention, what we can manage; based on the liability, what we can write; based on the controls and risk management, what we can do. I think I would look at our company, that we have ability to grow to that level. From a global perspective, on a country perspective also, the market, what we are, which is less than INR 300,000 income households, which currently touches the industry only by 40%, which means 60% of the households are yet to get into a formal market. The first formal market they would get into, microfinance. So if even -- if they convert another 40% in the next 5 years, I think it is still difficult until the market will double by that time. So -- and then also top this, there will be inflation impact on the loan side, there will be impact on the customer, vintage customer, higher loan per customer, all these factors together. So there will be enormous growth opportunity for the industry. And then after a certain set of penetration, there is 60% to 70%, it is more of a competition like what you are seeing today in the housing market. So I think then we need to really compete with your product, process, pricing, service, everything. I think we should be able to do that. I think, therefore, we believe that there's enormous opportunity for us to grow.

I see. So what are some of the big risks that you see for your company and as well as for the sector, that you can say with your imagination that these are some of the big risks for the industry as well as for the company?

We are not expecting any known risks actually. There could be unknown risk, which is unknown anyway. So known risk, if you see last 10 years or 15 years. The risks which are come to this industry are unknown actually. So whether it's COVID or demonetization, the second COVID, all are unknown. The industry has not lost because of the known risks. It's all very small, sometimes what you call a flood or a famine sometime. So these are small, concentrated places, so we are not seeing any big risks for this industry which is known and manageable. So unknown, sometimes unknown, but if you're able to manage right control, right risk management, right customer service, still we can manage better.

I think the CreditAccess group has set up some insurance business as well. So how are they going to use our distribution channel? Or how are we going to integrate with them? Or that is completely different business, and we have nothing to do with that?

It is a completely different business. It is, I can say, it's not even group company. It's a parallel sister company, I can say. But CreditAccess India Life is one more service provider for us like our other insurers. We have -- totally, we have 5 insurance providers. So one among them is CALI. So we keep supporting, we keep innovating the product for this set of customers because CALI is more of a low-income household focused insurance or a micro insurance segment. So there will be opportunity for us to co-work and innovate new products for this segment. So which is not an option with other mainstream insurance providers. So we will continue to work together to ensure the right product for the customers.

Do we have insurance for the -- if in case the borrowers are not able to repay, then the insurer will pay for a couple of installment or something like that?

No such insurance as of now for anybody in the country.

And the last question, what is the update on the income tax demand that we received from a year back or so?

Year back, it was closed, it is refunded to us already whatever we paid as advance. So only what is pending is what we received in the last month, which is about INR 46 crores. And we -- our auditors clearly said that there is no need of any provisioning or even contingent provisioning for this because we feel that there is an error in the assessment and we already made appeal, and we are sure that we will get this canceled, the demand gets canceled. It will take some time, but of course, it's a process, always. We'll process -- we will follow the process and we are pretty much confident that this is not the right demand.

So there was some big demand that is closed, right? There was a very huge demand. It is closed.

That was in 2021. And then it always -- it's reduced to INR 100 crores later, than it already canceled and removed. We have published that also. There's no demand -- no other demand on us.

[Operator Instructions] The next question is from the line of Hardik Shah from Goldman Sachs.

Congratulations on the good quarter, sir. My first question is on in the retail finance. Can you share some trends that you are seeing on the ground across products?

Ganesh here. So on the retail finance side, we are around 3% of our AUM is retail finance. Out of which, the unsecured individual loan portfolio, which has been extended for graduated customers, is at a rate of around 72%, and the other 28% comprises of secured loans, which includes loan against property, gold loan and affordable home loans. Our total RF book is at INR 708 crores, yes.

Okay. And what do you think is your right to win in these segments versus competition, Ganesh, and your confidence on scaling up?

Yes. So for the first product, which is your Unnati product, this is more of a graduating product for a certain profile of customers. And while -- like in microfinance, we do command a very low interest rate for this particular product. And it is selectively given to a certain set of customers centrally decided by us for various parameters, right? So -- and then it is reinforced or reassessed at ground for certain other surrogate, and only those customers we extend this product. So this is scalable, we believe, in the range of 5% to 7% of our overall AUM. And this portfolio has behaved consistently well. And since this is given to vintage customers, we believe that it increases customer loyalty for us and it benefits us over a period of time. With respect to loan against property. Once again, we command a far better interest rate compared to the market. And we also have a good sourcing channel, right? So we have roughly around 13,500 field officers going thousands of villages on a daily basis. And we are looking at a majority of this business, both home loan as well as loan against property, come from our existing customers. So that means you are going to significantly reduce your sourcing cost, try and cross-sell to a certain profile of customers who are looking at those, who are capable of repayment, who are able to establish cash flow. And hence, once again, you're retaining your customer better. So that means I will be able to do something like a mortgage business at a far lower cost than probably an HFC is doing today. So from a customer perspective, they continue their journey with us. They're comfortable in dealing with us. They're getting a product probably much cheaper than what they will get from any closest competitor in the market. And their journey continues with us. And if you take 2-wheeler, it's similar. So we don't take share from the dealer. So we don't ask for a market share from the dealer. So we source business and take it to the dealer, and hence, we don't incur sourcing cost. And because this is given only to our existing customers, it is something that is building a lot of loyalty for our customers these women are enabling a first new 2-wheeler in their home for their children, for their husbands, for themselves, which is quite unique, right? And this, again, is extended to them as a weekly loan, and it's quite successful at this point.

Understood. And one follow-up on this. How should we think about these products compressing yields on the overall book? Can you share yields on this?

Almost all the products except home loan are very close to our microfinance rates, which are still competitive compared to what is charged in the industry. Home loans alone, like we said earlier, it would take some time for us to build the portfolio. And hence, you will not see any immediate impact on the volume. And as the volume picks up, we will figure out methods of ensuring that it does not compress our yields.

Understood. One last bookkeeping question, if I may. Can you share your [indiscernible] to CREDAG customer in group lending?

It's 28% as of the end of financial year. As we keep growing in markets which are not new to us -- sorry, which are new to us, this ratio I think will slightly dip and then move up over a period of time.

[Operator Instructions] The next question is from the line of Sameer Bhise from JM Financial.

Congrats on a strong quarter. Just one question. What would be the interest rate differentiation between, say, a first-cycle loan and 3-year loan for, say, INR 75,000-plus ticket size customer?

So there is no differential on interest rates depending on the cycle. It's only the vintages. So less than 4 years and greater than 4 years. So there is a pricing benefit of 1% given to customers because the ECL model also predicts a much lower credit cost.

Okay. So that's only 1% across the vintage.

No. So less than 4 years, greater than 4 years. If the customer was greater than 4 years, they get a 1% benefit.

Maybe I will just clarify it further. I explained one thing. The pricing is based on 2 factors. One is vintage factor, one is the risk factor, the risk of geography factor. For example, in high geography, there's a higher pacing even for first year. But whereas if they're more than 4 years, then the low price, they will get the benefit almost 1.75% difference. So the lowest paid for the charge, lowest to high difference is about 1.75%.

Fair enough. This is helpful. And secondly, if I just look at the statewide GLP mix, Karnataka has dropped a bit, while the other 2 large states, Maharashtra and TN, the decline is slightly slow. Is it likely to remain that way, that while Karnataka share may drop while other -- and converge to the second-, third-largest states as the noncore states grow?

The change will not be significant, actually. It will remain, to a large extend between 18% to 19%, 18% to 20% by next 2 to 4 years. So it won't drop faster actually. So it will remain at that level.

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

Congratulations for the quarter. Sir, two questions, one I think you've partly answered or touched upon. In your PD LGD calculation, do you also take the vintage of the customer into account, apart from the district and all of that?

Yes.

Yes. Yes. We look at both factors, vintage and geography risk, both.

Okay. So as long as you're acquiring a lot of customers with low vintage and in new geographies or riskier geographies, then your blended credit costs will continue to go up. And as those markets mature, it should...

Yes, you are right. Absolutely right.

Right. And for geography, when does -- when do you take a geography or district to have become mature? I mean, does the credit costs have to show a lower trend, and that is when you say, "Okay, now this is matured." How does that happen? Or is it just any AUM -- ticket AUM size?

We recalibrate this district risk status every year. And then normally, we will take decision once a year or once in 2 years based on the continuous trend of high risk, low risk status. But every year, we will review and recalibrate the districts.

Understood. Understood. Got it. And just the second question on retail finance. What kind of, let's say, AUM are you looking at from a, let's say, next 1 year or 2 years' perspective? This year, of course, it's in a startup phase, so the growth is exponential. But what is your, let's say, target or milestone 2 years out?

If you look at what we declared in the last year during the Annual Meeting, we said we will reach about 10% to 15% of our book, and that's retail by FY '28. So currently, we have -- maybe every year, 2% to 3%, we would add to reach that. So which means we may have to reach around INR 7,000 crores, INR 8,000 crores by '28. And even if we convert our current 10% of customers into average INR 3 lakh in the midst all the products, so we will be more than INR 15,000 crores. Actually, we should -- we need only about 4% -- 6%, 7% customer conversion with the average about INR 2 lakh to INR 3 lakh portfolio. So that's why it is more of a in-house business to expand with very low all-in acquisition cost.

Got it. And just for these customers, you are cross-selling to -- or graduating from your group lending to individuals. What is the minimum criteria? As in what does the vintage have to be? Or what do you look at before you decide to cross sell?

So it's a combination of various factors, including the kind of district or state. So you will have vintage, you will have geography, you will have economic profile, you will have the type of business. So there are a set of targets that we've built. This is also something we consistently change.

A minimum of 3 years -- 2 years.

Minimum is 2 years. But they have to meet many other criteria.

But I think one strong differentiator, will be they have to be able to demonstrate cash flow better. And we also have to have multiple source of income. They should have segregated place for business, et cetera.

The next question is from the line of Shweta Daptardar from Elara Capital.

Congratulations on a good quarter. I have two questions. One is, what is the movement of average ticket sizes across the core markets?

So exact value, I don't know, Nilesh will share with you. It will always be higher movement in the core market because we will have a higher vintage customer in the core market. So exact number, maybe Nilesh will share with you separately. So newer market, the movement will be very low because it has to complete 2 years and then 4 years kind of thing. Therefore by default, our core market will have a highest movement.

Right. And sir, second question, what is the rate differential for your bank borrowings vis-a-vis our marginal cost of funding? Because since 2 quarters, we have been drawing down on our bank borrowing.

Yes. Actually, bank borrowing, normally, it's about 9.5% range. So this time, we are able to get a little more benefit for the bank, therefore it is about 9.3% to 9.4%. And international borrowing and the NCD this all will cost around 9.75% to 10%. So overall, about 25 bps more. But it's important to do the diversification because both funds are long-term funds, which will give a PLM benefit and also diversification benefit. So therefore, the overall cost increase by increasing the diversification is about 20, 30 bps only.

The next question is from the line of Mohit Jain from Tara Capital Partners.

Congratulations on a good set of numbers. I was just looking at the incremental -- the disbursement yield on the new portfolio. And that seems to be slightly higher than the portfolio yield over the last 4 quarters. And I think you said that the borrowing costs is expected to be flat. So in this situation, the NIM guidance that we are seeing for the next year is 20, 30 basis point lower than -- be fairly conservative? Or are we pushing the interest rates?

If you see our new disbursement rate is 21.8% in the Q3 and 21.4% in the Q4 because we actually reduced the pricing by 50 bps between December to Q4. So therefore, we expect, in the last next 1 to 2 years, there's impact of some 20 bps. So therefore, we projected a little lower.

Okay. Okay. And sir, just one more question. Sir, if I'm looking at one of the NBFCs, which also have a bigger NFI portfolio, they have resorted to monthly collections in the beginning of the month. And their argument is that saves the bandwidth of the manpower because they focus on the collection at the beginning of the month. And throughout the month, they can focus on the disbursement. I know, sir, we have always been a proponent of the weekly collection, weekly and fortnightly. So -- but have you evaluated ever that we move to the monthly one so that we'll have a better disbursement? So what's your view on the entire argument?

No. Actually, we -- I think we have done enough study internally about what is the best for the resiliency in this business. It's not just a collection for this month. And then we should not be looking at only based on the good time performance. We also need to look at the bad time performance, how this will differ between good time and bad time. So we have been experienced with the situation over the last 10, 15 years, and we strongly believe that given the option of -- to customers, rather than we deciding one way, either monthly, fortnightly, or weekly, instead of we give option to customers to decide what way they want to pay. So our -- we give the option to customers saying that whichever the period or the frequency you want to decide, you can decide when borrowing. So our customer, at least 58% to 60% customer have decided to go for a weekly. About 30% to 35% decided to go for fortnightly. And balance, monthly. So we have all 3 options to customers. Whereas we meet our customers every week, based on the choice of frequency that customer's chosen, they're paying. I think we thought these are best way. We have been operating in this model for the last [ 12 ] years, and we believe that this will work better in good times and bad times.

And also increasingly, since you are now doing various other things with the customer, it is better to meet them more and helps you with the relationship as well as offer other services. You're not restricting yourself just to one product today, you have multiple loan products in group loan itself. You have retail finance, you're offering products to the ecosystem.

Connectivity is much higher.

So you have a better relationship. I'm going to meet probably 4 times compared to people who are in monthly. But different school of thoughts, different behaviors. I think we should not criticize any of it. It's just time period, like what I say, good and bad times, that should determine.

The next question is from the line of [ Shreyans Jain ] from Electrum PMS.

All of my questions have been answered. I just have one clarification about this weekly, fortnightly and monthly model. So the customers are repaying, 60% are repaying weekly, and 30% fortnightly, 10% monthly. But are they coming to the center meeting every week or they're only coming when they know they have to pay their installment?

They will come every week, actually. They will come every week. Those who have opted for weekly will pay every week. Those who opted for fortnightly will pay only in alternate week. Those who opted for monthly will pay once in 4 weeks.

And what is the center attendance on the portfolio on the whole-country level?

It's about 65% to 75%. So there also, we have to note that normally about 12 weeks in a year, we allow them to take off, actually.

12 weeks.

Yes, that means almost 30% of the time they will not attend the meeting. It's a legitimate holiday available to them.

Ladies and gentlemen, as there are no further questions. I would now like to hand the conference over to Mr. Udaya Kumar Hebbar for closing comments.

So thank you so much for spending time in the evening, and it is quite late. Thank you for all your support. We hope we have responded to your queries. And if you have any more queries, our investor relations desk is always available to reach out. Thank you so much. Have a good evening.

Thank you. On behalf of Investec, that concludes this conference. Thank you for joining us. You may now disconnect your lines.