CREDITACCESS GRAMEEN LIMITED analyst Q&A

Sir, my question was firstly on this accelerated write-off during the quarter, so what was exactly the profile of these customers in terms of they being more than INR 2 Lakh ticket size or they being CreditAccess +3, CreditAccess +4, if you could just some color on that?

Sorry Shreepal, this is more based on the repayment profile because these customers are already more than 180 days and there is no repayment coming from them from a minimum 90 days. So, that is where we thought into a write-off. We have not gone through that kind of detail, which we can analyze and share with you privately after.

The second question was on X bucket collection efficiency trend from Sep’24 to Jan’25, if you could provide that will be really helpful?

For the month of December, it is 99.2%, I think it is more relevant what is latest, which we will continue and we expect January also continue in the same level. As of now, it is equal to 99.2% level. November and December , it was around 98.8 % and 98.7 % between October and November and for the month of December, 99.2%. January in a similar performance.

Is this available in your presentation actually?

In Slide #6, it is available. Actually, it is PAR 15+, but I think more or less you can see that , PAR 15+ is already available, that the reverse data is actually expected.

And sir, one last question was on the industry side, so with guardrail 2.0 being implemented from April 2025 and while we are seeing positive trends on a sset quality side and also on the disbursement trend that is highlighted in one of your slides in Jan, but how do you see things shaping up for us and for industry in FY26 in terms of trends on growth as well as on sustainable growth?

So, maybe I will be able to give trends for us actually because we need to see exactly for the industry after getting more balance sheets and publications by others. So, for us, we are clearly seeing visible improvement. Two, for new customers, largely we already implemented that guardrail. We feel that it is very important for new customers, though it is referred to 1st April .. And also we put that trend there in our presentation that even if it is implemented from 1st April, the implication is quite low because that set of cohort, already 84% are paying and our internal assessment clearly shows that we will be able to retain more than 80% of customers in that bucket also.

No, for the industry, it is difficult to say right now. We have actually estimated if you see the slide which talks about annual guidance, we said 18%-20% growth is estimated preliminarily by us for the next FY25-26 and for the current year, we are talking about 7%-8% growth.

Thank you. The next question is from the line of Neetu Tharakan from MFO. Please go ahead.

Mr. Ganesh, I didn't get the presentation, so maybe some of my questions is because, I don't have a physical, not able to see the PPT. But I am just curious, we see the trend of delinquencies across the industry, but when can we expect a normal figure coming in maybe like what Q1 of next year or it might take more than that?

I will take that. I think we did that initial remark, I think between Q4 this year and Q1 next year. Between that, we should be getting into a normal zone, Neetu.

I was just asking, what is the loan portfolio size now compared to last quarter and I heard that you are seeing a positive trend, so I am expecting it to grow in the next few quarters, but right now I don't have the figures for this quarter?

So, as of December, the AUM of the company stood at INR 24,810 Crore. As of January 20th, it is at INR 25,125 Crore. So, year-on-year, we have grown, quarter-on-quarter, we have degrown because of the accelerated write-off.

Thank you. The next question is from the line of Shreya Shivani from CLSA. Please go ahead.

Sir, my question is more around the business operations on the field. We had heard about loan officer attrition being high across MFI players and probably that number is north of 50% for the industry. Where would we stand? Can you help us understand what are the key concerns the loan officers have currently, and we had also picked up from some industry interaction that there is a peculiar case in the M FI industry that many of the loan officers who quit their jobs in the past couple of months, they actually quit and did not join any other formal sector. So, can you help us understand what is exactly going on at the branch level among the loan officers and at the operational level, not so much to do with what is happening with the customers over here?

Normally, in periods of elevated stress, the attrition does moderately go up, same for it, but however, for us di fferently, during this cycle we ha ve got roughly around 3 ,000 employees requesting to rejoin the CA Grameen. So, I think we will be able to manage even if the attrition goes up slightly, but right now it is not a big challenge for us.

Sir, is there attrition rate higher than 50% or lower than 50%? Is there some number you can?

And sir, this is not just specifically about you, but if you could just help us understand, yes, in times of stress, the loan officers do exit, but is there a trend where they exit and they do not take up jobs in the formal market because we picked up this in some of our conversation with some other industry experts?

It is a combination. Some of them do, some of them don't. So, it can be probably equal in number. A lot of young guys and girls who join, test the waters, and if it doesn't suit them, they leave the industry and if it doesn't suit the organisation, they go with some other company.

And sir, in times of stress, does the attrition rate pick up more in loan officers who have joined within the past one year or with more than 2 years or 3 years or was there a change in the way hiring was done that did the industry hire people with lower expertise or something like that? Had something changed in the last 2 years that led to much worse behavior in attrition trends currently or because of this normal trend that has gone on?

It is a normal phenomenon under stress. Like I said, a lot of pressure will come in. They will not be able to kind of handle the situation, but once an employee normally crosses 6 months or 1 year, they don't have all these issues.

And sir, my last question, just one follow -up over here. Have we changed any requirements in our hiring process for these loan officers? Have we tweaked it upside or downside?

So, you would know that we always hire only fresher s and we stick to our strategy and there is no tweak in any of our processes.

Thank you. The next question is from the line of Sanket Chhed a from DAM Capital. Please go ahead.

Sir, my question was on our guidance, particularly this year we have been stretched on our credit cost guidance as we move quarter after quarter. And now, as there was earlier question that from April, the guardrail 2.0 will be in the effect and we have different shares in terms of CreditAccess +3, 3+ and 4+, so w hat makes us so confident to give this guidance of 18 %-20% growth plus ROAs also, almost in the range of normalized levels and ROE while it has been a turbulent last few months. November was just the peak. After that it is just 1/2 month that we are seeing some improvement, but we are yet to solidify those trends. So, what gives us that confidence to be so early in terms of guiding mostly a normalized year this is FY26?

So, if you look at our earlier guidance about p eaking out and how we will normalize, we are more or less there. We have got a 4-week delay, but otherwise we are already seeing that pattern emerged more than December as well as January. And we a re confident that we should sustain it during the year, right. And like we said earlier, all our internal analysis, we have done a full PR of our customers, etc. Basis our internal analysis, it gives us enough confidence that the next year growth also irrespective of the new guardrails coming in, we should be able to maintain and we have strategies around it.

Sanket, I will add here. So, see one thing is that if you see always, a majority of our growth has come from customer retention, historically 60% growth has been from retention of customers. So, as we have shared data, we have a fair visibility on retaining the customer cohorts who have say more 3 or 4 loans because a majority, more than 80% of those customers have been promptly paying. So, as a result of that, there has been a normative deleveraging which has happened over last 6 months and which will continue to happen over next 3-4 months as well. So, in a normative fashion, they will become eligible to avail future look. So, that is where it gives us visibility to retain our customers and second is the addition of new customers. In the month of December, we have added 70,000 customers, the current run rate in Jan suggests that we will cross 80,000 and in a normal period, we have been typically adding around 1 Lakh customers a month. So, adding 1 Lakh customers a month, next year should be achievable for us, which will give us a high single -digit growth in borrower base and overall microfinance portfolio maybe around 15%-16% and on top of it, we will also be getting growth from retail finance which has been demonstrating healthy growth. As you see over last one year , retail finance had grown from 2%-5% and lastly it is behaving very well from the asset quality point of view as well and it is a step towards retaining our high vintage customers. So, a combination of all these factors gives us the confidence to deliver that growth in the next year.

And in our ROA guidance, what have we assumed in terms of the credit cost for next year?

Sanket, like if you see our guidance this year, we had earlier guided around 5%, now we have revised it to 6.7%-6.9%. So, basically 1.2%-1.4% delta, it is on account of a month of delay in the improvement in the reversal in the delinquency cycle. So, that is where 1% has added to this year's credit cost. And now, as we said, we are going to take accelerated write-off over Q3 FY25, Q4 FY25, and Q1 FY26. So, that will help us to absorb more than 80% of the current year’s impact in the current year itself. From that perspective, maybe in a normative year, if our credit cost is around say 2.0%-2.5%, on top of it, may be another 75-100 bps might get added into the next financial year. So, that is where for next financial year , as of now , the estimation is we should have a credit cost of 3%-3.5% which will give us an ROA of 4.2%-4.5%.

From an ROA point of view, Sanket, in the current year, our ROA is less by 3%, a total of 3% ROA, it will take this year and 1% RO A, it will take next year. So, net-net overall the delta of ROA hit to us because of this event in both years, together is only 4%, 3% this year that means we will be delivering almost 4.0% next year, we will be able to deliver ROA of about 4.5%.

Sir, my first question is, sir, while the new PAR addition is now lessening, any improvement is also seen in resolutions across SMA buckets, can you comment on that? Is ample recovery also improving along with the new PAR getting less added, are you seeing improvements there also in the bucket reduced with an NPA?

I think it is an improvement in both sides. One is a reduction in new PAR accrual, second is the improvement in the forward. Forward is actually reducing and the increase in the momentum so all three are combination actually, Rajiv. You see, almost 40% of Stage-2 bucket already in a partly payment stage and the reduction is rich as X b ucket. The collection efficiency is more than 99% now.

So, sir, what do you attribute this improvement to collection across buckets because as I see, you haven't added employee or collection team like so many peers end because your employee base is actually flat. So, what do you attribute this improvement in collections? Is it attributed to any improvement in center m eeting attendance, better customer reach out and how have you done it?

No, I think we anticipate early also it is more of a transient nature and I think we estimated that it will p eak out and start coming down the new PAR, I think it happened in October from November. Several days, we are also able to deploy more people, experience ones if there are pockets of collections to work on that and then the underwriting quality is changed for at least new customers and new disbursements. So, all these are a combination of things and then , we didn't have so much of attrition also. If you compare the attrition, it is not a very large deviation from normal. That also helped us . All these together in a combination effort, I think where the customer was defaulting largely have already defaulted. I think there is no more new customer defaulting now, very less new customer defaulting. That also is the trend we are seeing now. So, overall, there is a good improvement in terms of both sides, improvement in terms of collection, improvement in terms of disbursement, improvement in terms of lesser forward flows, all these are happening together.

In times of such , all our Control Teams also support us with collections. I also wanted to add, Rajiv, as a demonstration to the field te am, almost all employees in the management , so the entire top management a dopted zones and all of us have traveled extensively over the last few months, including meeting PAR customer collection more towards the demonstration to the field force, right. So, everybody is in the front. I think the biggest difference in such scenarios will be your field supervisors, your seniors getting involved in solving the trouble. And like I said, we also have an additional workforce with a very high number of control teams. So, they also contribute in such times of stress to support the field teams to come back to normalcy. And you would know about our people strategy, how we play on hiring only freshers and retaining them. A lot of our seniors have seen multiple cycles, right . They also know exactly how to navigate the scenarios. It takes some patience and hard work, but it is something that we have done well so far.

Just one last thing, any revision in lending rates that we are evaluating in the light of increased credit cost and maybe even OPEX because when I look at your NIM guidance for the whole year, it implies that in Q4, your NIM has to improve by 30-40 basis points over Q3. So, what will drive this NIM improvement because you will still have higher interest reversal s in this quarter, so what we cover up for it? Have we taken any rate hikes?

So, Rajiv, our pricing policy is very clear and based on certain data actually. Based on the variation whether in OPEX cost or the credit cost or the borrowing cost, it will change to some extent every quarter being reviewed. I think there is a small review that happened in Q3 where certain basis points the change was there. Similarly, as and when there is a change, every quarter, there will be change. So, there may be a small variation because of that, but because of the higher interest reversals, the NIM is actually moderated back to the same level.

And to a certain extent, the pricing policy does not change much because it is a long-term average that you take . You take a few quarters' average and keep repricing as and when required . So, there is Board approved policy for this and every quarter if the credit cost increases for the next 4 to 8 quarters, then it will gradually get passed. So, right now, we don't see a big movement in pricing, but we should come back to normal sequentially.

Thank you. The next question is from the line of Renish Bhuva from ICICI Securities. Please go ahead.

Sir, just two questions from my side, one on this PAR team plus accretion, so when we look at the state wise numbers, of course there is an improvement across states except Tamil Nadu. And when we look at the Karnataka specifically, though there is an improvement in January, but there are lot of news flows which sort of keeps on coming over last month or so wherein some district, some pockets, some external events are happening. So, how confident we are that the kind of improvement we have seen in Karnataka maybe over the last 2 months, from a peak of 1%, we are now down to 60 basis points that will sustain?

You are right, the re is some negative news around microfinance in Karnataka, but this was actually happening for the last 2 months, majority was in Gulbarga and Belgaum districts. There are some impacts we saw in the last month there also . Despite that we are able to show better performance in Karnataka, but recently there is more news in the media. A lot of microfinance news in recent days, there have been few local third -party interventions, particularly at what I said, Gulbarga, Belgaum, Tumkur, Ramnagar districts. Our members and industry bodies have addressed and appreciated to the administration of respect ive district as well as Government of Karnataka on the governance of microfinance, malpractices, code of conduct , grievance redressal system, regulatory entities like RBI and SRO 's governing th ese REs , t hese are appraised to them and how these are handled well . Based on various publications, I think the Chief Minister of Karnataka also called a meeting tomorrow of stakeholders including the members and including RBI. I think this will serve a platform to discuss the role of microfinance institutions contributing to the state economy, how the microfinance is actually running following the fair practice code, the regulations and governance. The RBI and SROs can explain what they were supporting and handling. I think this will help us to build a strong connect. The outcome of this meeting will be productive and a step forward for us. I t hink at the ground level, collections are going on smoothly, no problem. But ye s, there is news, particularly what you observed from a non - regulated entity. There are one of the issue difficult to identify and inform, but whether you are informing the police also. I think we will be able to highlight these issues to the administration and the government tomorrow probably. We definitely believe that this will be a good step forward toward resolving this issue.

At this point in time, is it fair to assume that you don't foresee any risk to the sort of improvement what we are expecting in Karnataka?

No, we don’t see, actually, even if you see the worst scenario in October, November, Kartanaka X bucket collection is more than 98.5%, right. So, right now, it is almost 99.4%. Therefore, we don’t see any major worries in Karnataka.

And s ir, again, the second question is related to that. In Tamil Nadu, though there has been improvement in December, but then again, there is a very sharp deterioration in Jan, of course, we did mention that these are transitory and maybe things should improve in Feb, but is there any lead indicators in place wherein sort of we are expecting things to bounce back in February?

Tamil Nadu, again it is more of a lag impact of December month ’s rain and cyclone. There is a small increase in PAR 15+. PAR 0 is still normal right now. PAR 15+, there is a lag impact, but most of the customers are partly paying, not able to pay all four installments. That is why it is still there. So, therefore we don't see this is a major issue in Tamil Nadu also. It is a temporary, maybe this month, by month’s end, more things will sort out, so therefore we don't see too much of issues in Tamil Nadu.

And sir, my last question again on the guidance side. So, now, when we look at our FY25 performance versus the guidance, couple of time we have been forced to change our guidance. Of course, I do understand because of the ongoing situation where in industry, so dynamic, every 2 months things might change and again hence, what sort of confidence you guys have that whatever guidance we are sharing for FY26 that will hold true?

This is an extraordinary year, Renish. Otherwise, probably there would be chances to change the guidance actually, but the most important point is to inform you upfront about what we can do . Actually, that is what we are doing also. When we compared to the revised guidelines, we are actually there only. Even we are saying that we are achieving between 7 %-8% despite we are taking the write-off, external write-off. Therefore we are actually not changing the guidance actually. By the way, changing guidance is on account of credit cost. It is again because of the extraordinary write-off we do plus the small delay in our estimation. We estimated the peak out in September, but unfortunately it was October , it has delayed by one month that costed us about 100 bps additional cost. Otherwise, we remain in the guidance. That is not too much change in the guidance. And for next year guidance , we have calculated carefully , it is the preliminary estimation, but largely we should be there. So, that is what we estimate an internal confidence we have against that. But definite ly we will come back with the full guidance in the month of May after the annual performance.

Mainly, let me put it this way, so what are the risk s which can lead to sort of we revising this guidance or we revisited this guidance. Internally, do you foresee any risk to this guidance or maybe the parameters which you track very closely, of course , collection is one number, but anything apart from this in terms of, let us say, the districts which are in stress currently because of external events or maybe some particular event happening in some state , etc. What are the key things which you guys would be watching very carefully to ensure that guidance remains?

I think we are clearly seeing the improvement which is visible for last I can say 2 months now, from mid-November to almost mid-January. We have seen the change in terms of the reversal of trend , in terms of delinquency, in terms of disbursement and in terms of new customer additions. We are able to see unless drastic change in this trend reversing that. Otherwise, we don't see any other reason for failure in the guidance.

Thank you. The next question is from the line of Bhavik Dave from Nippon Mutual Fund. Please go ahead.

Sir, just a couple of points, right, one is again on the guidance bit, we missed the guidance like 2-3 quarters in a row. I think it will be good to maybe get out for the end of the year before giving guidance on growth because my question comes back to growth, right and when I look at your numbers, you have been adding 1,50,000 customers per quarter. You mentioned that you will add 1,00,000 per month and even if you do that and when I do the math , 15%-20% growth seems to be bit of a stretch an d in that context it would be great if we could just maybe give us a sense that why will customers after the three guardrails stick to us versus going to someone else who might give a higher loan? Considering the customer segment we have seen as we have very similarly for all lenders, right , it is not that customers are paid X lender and not like I defaulted to the Y , unfortunately, over-leverage borrowers have default behaved in a similar way. Second is on the retail finance as well , majority mind our exis ting customers and given upgraded them to retail financing. So, a large part of our customers will also be upgraded in that sense, right? The 70%-80% of the customers that pay back after this washout, it seems that the customer that we will be able to lend in the existing format and also the new customer addition seems to be a bit of a stretch is my number that I am working around with. So, if you would explain us how this 15%-20% growth will come through because I am able to do a little math in terms of the customer and the ticket size that we are working with?

I think broadly what we have guided the growth will come from both retail and microfinance. So, we have always been saying that as we grow bigger and bigger, the growth rate of microfinance will slow down. You have a lready seen that our share of retail has moved from 2%-5%. So, with every year moving by, the contribution of retail will continue to grow. And again, we always stated that microfinance will be an entry point for the customers and only a certain profile of customers will be abl e to upgrade. Not necessarily that everybody will move toward retail. People will have to demonstrate certain history, a certain amount of cash flow to demonstrate for them to move towards reta il. So, I think there is enough room in whichever market we are. If you look at any of our non -bond markets, we are still very small and I think there is enough room for us to grow both in microfinance and retail, but increasingly the growth in microfinance will t ake a step down and retail will catch up. So, with a combination, we are looking at delivering this kind of growth.

And also, sir, when I look at your Bihar PAR here, I see that outside of 3-4 strong states of us, our performance has been quite poor because we have heard other lenders talking about we are recovering quite well for them. In that context , ho w has our experience been both in microfinance and retail when it comes to the non -core geographies like in a sense, Karnataka, Maharashtra, Tamil Nadu seems to be our strong point. Apart from that, how has our experience been in terms of retail and MFI in the cycle, has it been far worse or what are you thinking about on those line?

So, today, we operate retail only in our core geo graphies which is Tamil Nadu, Karnataka, Maharashtra, and Madhya Pradesh. We are not there in other geographies because our strategy around retail is to bring in customers, season them to a certain extent and offer retail finance. So, probably, we will take a few more years to reach our non-core geographies. Once we become a certain size in each of these geographies is when we will overhaul our retail finance strategy. With respect to our retail finance portfolio, it has been absolutely good even in the first time of stress. We are going through a very high credit cycle. However, the retail finance book has held really good asset quality . I think PAR 30+ as of December, even for un secured graduated business loans is in the range of 70 basis points. So, this is unsecured, secured even much lower.

Unfortunately, sir, what you see is most of these MFI customers have other retail loans and that is what is causing the overall dip just like other industries are just looking at maybe borrowers who are three or four lenders to them. But other than that they have other retail loans and that cohort has done even worse than microfinance customers. So, that is the reason I am trying to maybe drill down a little more?

One comment on your last statement. So, even though our presentation, we iterated and even the numbers that in our case, while it could be some numbers published by different agencies. In our case, what we see is that there is no difference in credit costs between customers who are only microfinance customers as well as microfinance customers with retail overlap. In microfinance customer’s case, the PAR is around 6% and in those customers with retail finance, PAR is around 7%.

You can refer to slide 11 for this.

And also in all our retail finance journey, we have consistently reiterated, we are not in a hurry to grow. We will test it, we pilot it each of the businesses for enough time, tested results against these pilots, and only when they scale up. So, there is a clear strategy around whom we will offer retail being secured or unsecured . And what is the marketplace , what is the space at which we want to grow, so it will be a natural progression over a period of time and we are very mindful about keeping this quality intact.

Thank you. The next question is from the line of Arjun Bhatia from Bowhead. Please go ahead.

Sir, we wanted to understand if by any chance get disclosures about loan repayment schedules off your existing customers to other microfinance companies as well?

Disclosed loan repayment?

A customer, let us say, he has got l oan with three other people or four other people , do they disclose with you the repayment schedules like when is the loan getting foreclosed from the other?

It is available in the bureau reports.

So, you have entire access to the data?

Yes. So, we may not have the entire repayment schedule, but we will know when was the loan disburse and when is the closure.

And have you used that modeling somewhere in your analysis?

Yes, you have to use it for even FOIR calculation, right. So, the regulations states that you have to calculate their overall liability, what is the repayment they make as a family every month and then derive FOIR out of it.

So, broadly what you are saying is that you have mapped already which customer is going to , loan is getting foreclosed? And therefore, how many loans will get foreclosed at your end versus others? What I understand from guardrails 1 and 2 is that whoever loan it has to be repaired, first that loan cannot be renewed is more than 5, t hen it has to come down to 4 and in the second guardrail if it is 4, it has to come down to 3 and so forth. Is my understanding correct?

So, in that sense, I wanted to understand that how do you get the comfort that whosoever is paying you now, after the guardrail 2, he gets choked in terms of further supply, he will end up paying to you, because he is not going to get a repayment. Just wanted to get some perspective on that? And you model all that in your forecast for growth, all of these like whole get foreclosed first at your end versus the other because wherever it gets foreclosed first, the customer can't come again?

Yes. So, we have taken everything into our model and that is also the reason for our confidence. And again, taking a look is the customer’s choice. Not everybody closing the loan asks for a second loan. So, that is when you need to continuously run for new customer acquisition also, right. So, it takes care of many things for you.

Secondly, sir, I wanted to understand, based on discussion with some of the players, we are told that there is a possibility of new regulations on capping the NIMs directly and indirectly through various ways. Additional top up loans, one could give, we understand, we give some top up loans, what are your thoughts on that and do you expect any further regulation at some point of time in the next 6 months, 1 year which will lead to further consolidation in the industry?

Right now, we are not privy to any such information. I don't think we should speculate about it.

Thank you. The next question is from the line of Abhijit Tibre wal from Motilal Oswal. Please go ahead.

So, the first thing that I want ed to understand is, what are the covenants that we have in place from our lenders? What is the GNPA or GS3 threshold in our covenants and what is the threshold in the quarterly profitability?

Abhijit, Nilesh here. So, there is a wide range of covenants. We have more than 70 lenders and covenant levels will obviously keep varying from lender to lender. So, we have a good amount of comfort, typically in the industry you will see the GNPA crossing say 4%-5%, then there can be some kind of a challenge. So, as of now, we have no such issues.

Basically, whatever thresholds are there on the NPA or GS3? We are at least much lower than whatever those thresholds are, is it?

Yes, as of now, we have not faced any issues there. That is the reason why the access to funding is continuous for us and as we reported even in the 3rd Quarter, we have drawn close to INR 4,000 Crore from more than 17 institutions and from our borrowing cost, there has not been any change. Our marginal borrowing cost continues to stay at 9 .4%. And at the same time, there is continued confidence what we are able to maintain with our lending partners.

And also remind us while we took accelerated write-offs in this quarter and we spoke about taking some accelerated write-offs in the next quarter as well. What is that usual write-off? What is he just trying to understand this accretion that we are seeing in GS3 right now, at what point in time to start to be written-off?

So, in our normal policy, we write -off after 270 days. And during even COVID, we had taken some accelerated write -offs. Similarly, now we have taken write -off for different portfolios which are not being paying for the last 3 months and 180+ dpd.

It is actually less than 1% of the portfolio we took the write-off, which has not paid for the last 3 months and already crossed 180 days.

And s ir, lastly, just trying to understand, while we did speak about Karnataka, whatever is happening there and you also spoke about that despite all that we have managed a good collection efficiency, let us say, when you look at your X bucket collection efficiency on a national basis and when you look at Karnataka, how lower is Karnataka from the national average?

Abhijit, if you see Karnataka, it is around 20 bps better than the overall bucket.

And just one last question then, of all the collections that are happening, what proportion is kind of getting collected in a central meeting and what proportion of collections are happening through door knocks?

Predominantly, it will continue to happen in centr e locations. Sometimes what happens if the field officer is delayed for the second meeting when the customers don't wait. So, there they go and collect from the Kendra leaders. Some of them if they don't pay that is when we go for door- to-door collections, but that is supported with massive teams, including control teams, field supervisor. Probably as an indication may be less than 2%.

Thank you. The next question is from the line of Parth from Nomura. Please go ahead.

Sir, on the delinquency trends, you mentioned that the y are transitory in nature, and you have been seeing reversing trends since November and December , mid-November. So, on this, sir what has changed on the ground that gives you confidence that this will stay? This is just transitory in nature in the incremental reversals and the delinquencies would sustain going ahead? And sir, my next part, the new borrower addition rate has seen a significant uptake during the month. What has changed on your onbo arding processes or your underwriting processes, which gives us confidence that these borrowers would be better in nature than the previous borrowers?

On the first question, I think the whole presentation demonstrates with sufficient data as to why we think this is sustainable. However, they have to see how it goes for Jan, Feb, March, right . So, we are reasonably confident we peaked out, we are showing as on date whatever is happening in December and Jan. W e have also demonstrated that overall , we are able to see leverage coming down and we have also demonstrated in our presentation how the previous cycles where and how much time it comes back. So, you would see in our presentation what we did in COVID, what we did in Demonetisation. So, overall, roughly 3 quarters, disruption is what happens for us to come back to normalcy, and we believe it is similar in nature now that we started showing performance in the last 2 months. New customer additions, we are looking mindful. We have tightened like, say, the g uardrails, we have been already doing voter ID validation , the field verification processes have been strengthened in geographies where we are seeing certain higher delinquency, we moved house verification one level up. We also used quality control teams to vet in customers joining us new in all geographies where we have elevated credit cost, it is a continuous process. And in some places, we have also made it stringent with guardrails where w e limited the overall maximum outstanding to INR 1.50 Lakh and also the maximum number of lenders to be not more than 2 and 1 in different locations. So, basis that district behavior, we follow a di strict model. Depending on the slippages in each district we keep moderating our operating procedures.

Our business rule engine gives us the flexibility to adopt or differentiated underwriting process at the district level.

And just one more question, so I understand Bihar and UP are your core geographies, but some of your peers have highlighted some collection efficiency issues in those two geographies. And while on the PAR accretion part, we have shown some impro vement in the December and Jan 25 numbers, so what are you doing differently in Bihar and UP where you have seen better collection efficiencies are improving trends there?

See, I think like you said, different people have different geographies depending on which district they are present, which part of the state they are presenting. However, Bihar also has demonstrated that we have significantly come back in the last 2 months. So, we hope this will continue in the next 2 months, in the quarter also.

Just to add here, so that whatever increase we have seen in July, August, September, as we said earlier, it was kind of an accelerated increase because the underwriting got tighter in the industry and now because of the deleveraging happening , we are of the view that customers who were not in position to honor their repayments or wherever they were not able to handle multiple loans or they were old leverage, those customers have kind of fall into higher buckets PAR. But the other customers, they have continued to repay on a prompt basis. So, as we have provided in our presentation, even customers with multiple loans or higher slippage, 80% of those customers have been able to pay us on a prompt basis. Overall UP and Bihar, it is close to 6%-6.5% of the overall portfolio. So, even their larger delinquencies kind of have played out in the month of October, November and that is where we are seeing the new delinquency addition rate slowing down at a faster pace in December and January. So, our approach has been consistent across all the markets, but yes, different markets are behaving or kind of reversing at a different phase. If you see , in Maharashtra , in Madhya Pradesh, our portfolio quality has been much superior compared to existing.

Thank you. The next question is from the line of Kamal Mulchandani from Investec Capital Services. Please go ahead.

Firstly, like if you could just let us know that if you are facing some attrition at the branch manager level as well and if yes, like what would be that number for us?

No, not very different from the earlier time. We have been range bound and we remain there.

What is that number?

Around 30%. There is no extra attrition we saw in this period, and we saw little whatever changes, slightly variation only in the loan officer level where we always keep recruiting and having sufficient backup actually. So, that is why the attrition is quite stable.

And again Branch Manager and above because we only do internal promotions, there is always a bench which is ready to become Branch Manager in case attrition goes up. And also like I told earlier, because it is a tough time, people who join different institutions or have left CreditAccess Grameen placed the request to join back . We roughly have around 2 ,900 requests to join back CreditAccess.

Also, I would like to know that what is our net forward flow rate from 1 to 30 and to 31 to 60 bucket and what was it earlier versus the current year?

Flow forward temporarily has increased.

Typically in normative times, we have seen that around 50%-60% of customers in 0 to 30 bucket, they do roll back and balance roll forward post which as you see our provisioning rates, typically we provide 70% in the Stage-3, which means that roll rates are in that range, around that in the normative times. During the last 4-5 months, we had seen the roll forwards were higher by 10%- 15%, but now it is kind of again reverting to the earlier in month of Jan uary. So, we are seeing a reversion in the roll-forward rates.

And lastly, if you could just help me what were the interest reversals during this quarter and during Q2?

In Q2, we had around 35 Crore of reversal and now in Q3, we have around INR 75 Crore of reversal. So, in this around close to INR 25 Crore is on account of the write -off, overall, INR 376 Crore whatever we have written -off and balance INR 50 Crore is on account of Stage-3 accretion.

Thank you. The next question is from the line of Ash lesh Sonje from Kotak Securities. Please go ahead.

One qualitative question from my side and apologies if you have answered this in some form before because I joined a bit late. Sir, i n the past few months, we heard of several lenders narrating unpleasant anecdotes , f or example, instances when several lenders were queuing outside the doors of delinquent borrowers to collect their respective installments and that resulting intentions at the local level , are those kind of issues now largely sorted declining meaningfully?

Our door-to-door collection is probably under 2%. And I think because we have a weekly model, rural model and we also have a larger understanding of how to navigate those issues. We don't see much of an issue, the point that you are making right now.

Thank you. Ladies and gentlemen, that was the last question for today. I would now like to hand the conference over to the management for the closing comments.

I take this opportunity to thank all of you for your patience and time in joining this call and for very interesting questions. I hope we have answered all your questions and in case there are any further questions, do write to us to reach out to us, we will be happy to clarify them. Thank you so much.

Thank you. Ladies and gentlemen, on behalf of HDFC Securities, that concludes this conference. Thank you for joining us and you may now disconnect your lines.