Arman Financial Services Limited (531179) Earnings Call Transcript & Summary

Ladies and gentlemen, welcome to the Q3 FY '21 Results Call of Arman Financial Services, hosted by Emkay Global Financial Services. We have with us today Mr. Jayendra Patel, Vice Chairman and MD; Mr. Aalok Patel, Joint MD; Mr. Vivek Modi, Group CFO. [Operator Instructions] Please note that this conference is being recorded. I would now like to hand the conference over to Mr. Jignesh Shial, research analyst of Emkay Global. Thank you, and over to you.

Yes. Thanks a lot. And on behalf of Emkay Global, I would like to thank the management of Arman Financial for giving us this opportunity to hosting the call. Along with us, we have Mr. Jayendrabhai Patel, Aalok Patel and Vivek Modi, the top management of Arman Financial. So I'll hand it over to Mr. Jayendrabhai Patel for the opening remarks, and then we should open for questions and answers. Over to you, sir.

Thank you, Jignesh bhai. Good evening, everyone. Thank you for all taking the time to discuss our financial performance for the third quarter and 9 months of FY '21. We have issued a detailed press release and investor presentation for the past quarter. Hopefully, all of you have had a chance to review it. As always, I hope that all of you and your loved ones are healthy and doing well in these unprecedented times. Indian economy is picking up steam and recovering from the COVID-19 pandemic over the last few months. Generally speaking, we are seeing income-generating activities increasing in the rural areas and hope to see them resume to pre-COVID levels as the vaccine distribution picks up. We are also encouraged by RBI's announcements to harmonize the regulatory frameworks for various regulatory lenders in the microfinance space, which would stop the regulatory arbitrage among the non-NBFC MFIs, microfinance practitioners and create a more robust industry to prevent overleveraging by the microfinance clients and strengthen client protection principles. This is something that the NBFC MFIs have been requesting for sometime now and is a welcome move. On the business front, the consolidated collection efficiency has picked up significantly from 87% in September 2020 to 92% in December 2020 and 93% in January 2021. This continued progress on the repayment front on a month-to-month basis is encouraging and shows both the resilience and the positive intentions of our customers. I will now give a brief overview of our financial performance for the third quarter and post that, touch upon collections, liquidity and disbursement in more detail. At the end of the third quarter, our consolidated loan book stood at INR 720 crores, lower by 13% year-over-year as higher repayment rates combined with lower disbursement in 9 months led to a rundown in the loan book. As you all know, being a very conservative lender, we had essentially stopped disbursement during the moratorium period and slowly started disbursement from August. In the Microfinance division, we initially focused on renewal loans for our existing clients base who had completed their previous loans. This is because a lot of the credit bureau data was stable -- was stale, sorry, in the MSME and Two-Wheeler segments. We have started disbursing to all customers, although with a more stringent underwriting process. As a policy, we did not indulge in any top up or net of loans in any of our products to prevent evergreening of our portfolio. Loan disbursement for the quarter stood at INR 182 crores. The pace of disbursement is expected to reach pre-COVID level this Q1 FY '22 onwards. The portfolio decline has stabilized and has started increasing. Our Microfinance and MSME portfolio stood at INR 549 crores and INR 117 crores, respectively, at the end of Q3, lower by 7% and 12% year-on-year. In the urban Two-Wheeler segment, our performance was impacted by the decline in two-wheeler sales in the last 1 year, given the challenging economic environment. However, our newly launched rural two-wheeler product has demonstrated relatively better performance, reporting a year-on-year assets under management growth of 20% to reach INR 11 crores at the end of December. The rural two-wheeler book now constitutes approximately 20% of our total two-wheeler portfolio. Gross total income declined 16% to INR 48 crores due to declined portfolio and softer yields on post-COVID disbursement coupled with lower booking of processing fees on account of lower disbursement. Similarly, our net total income decreased by 14% year-on-year at INR 31 crores during Q3, led by rundown in the loan book. The company continuous -- the company's continuous efforts on rationalizing operating expenses bore fruit as OpEx declined by 5% on year-on-year to INR 13 crores in Q3, while our cost-to-income ratio stood at 41.3% in Q3 FY '21. In keeping with our conservative approach, we strengthened our provisioning coverage by prudently recognizing provisions of INR 15 crores during the third quarter. Further, we also took an aggressive write-off of INR 7.04 crores in Q3 to help reduce the NPA burden of pre-COVID doubtful assets in the future. Including the additional provisions recognized during the second quarter, cumulative total provisions at the end of Q3 stood at INR 52 crores at the consolidated level, covering approximately 7.2% of our loan book. At the stand-alone level, cumulative total provision stood at INR 18 crores at the end of December '20, covering 10.6% of the total AUM. Strengthening our provision coverage should help us deal with any impairment on account of COVID in the future. As a result of the higher provisioning, our net profit stood lower at INR 2.9 crores for the third quarter. Adjusted for the higher provisioning on account of COVID-induced disruption, profit after tax would have been substantially higher. Our gross NPA and net NPA has continued to remain low and steady at 0.7% and 0.0%, respectively. The NPA numbers are after considering no change in NPA recognition after 31st August in accordance with honorable Supreme Court's order. In the Supreme Court order did not -- if the Supreme Court order did not exist, gross NPA and net NPA would have been 4.1% and 0.6%, respectively. In terms of capitalization, we remain adequately capitalized, with consolidated debt-to-equity ratio of 3.1x. Liquidity-wise, we are in a very comfortable position right now. As on December 31, we had cash reserves of approximately INR 127 crores, including the undrawn CC limits. Liquidity position has improved driven by the pickup in collections and in the incremental debt capital raised since the start of April. Further, we have also raised INR 64 crores at attractive rates since September '20 to bolster our liquidity position. As per the RBI guidelines, the company has provided relief to distressed customers, allowing a onetime restructure. Thus, restructuring was provided to willing MFIs and MSME customers who showed genuine distress, and an EMI deduction was feasible by increasing tenure. No moratorium or additional top-up loans were provided. As of December 31, 2021, 5.6% of our loans, which 3.4% were 90-plus overdue, have been restructured. Now to give you a more granular breakdown on collections. In the Microfinance segment, particularly, the improvement in collections have been very encouraging as the repayment rates have been jumped from 84% in September to 91% in December and 92% in January. In MSME, the repayment rates improved marginally from 93% in September to 95% in December and in January. While the -- in the two-wheeler segment, the repayment rates stood stable at 96% in December and 97% in January. With the recovery in the economy and pickup in repayment rates, we now feel comfortable to gradually and carefully scale up our disbursement. Hence, going forward, we expect pre-COVID growth in disbursement from Q1 FY '22 onwards. Further, to aid disbursement, the company has expanded its footprints with decision to open Haryana with 10 branches in January '21 -- 10 branches in Haryana in January '21. Also added 3 more branches in Rajasthan, with a plan to open further 10 branches before the fiscal year-end. Finally, to conclude, I would like to express my gratitude to all our stakeholders for their continued support during these difficult times. A special note of appreciation for the company's field staff and also other team members whose perseverance and untiring efforts are all the sole reason why we have been able to report steady improvements in repayment range -- rate since June. Overall, our confidence to successfully navigate our way through this storm gets stronger every month and every quarter. I would now request the operator to open the floor for questions and answers. Thank you all.

[Operator Instructions] The first question is from the line of Amit Mantri from 2Point2 Capital.

Sir, just wanted to understand this GNPA that you reported of 4.1%. In this, the restructured book is also included, the 3.4%, which is 90-plus overdue? Or is that separate?

That would be separate.

So if there had been no restructuring, then the GNPA number would have been 7.5%, is it?

7.2% or something like that, I believe. 7 -- a little over 7%.

Okay. So 7 -- GNP, including the restructuring is 7.2%? Okay.

Something around that line. It's a little over 7%. Whether it's 7.2% or 7.3%, I don't recall at this point.

Okay. And what percent of the loan book has not paid a single EMI?

So in Microfinance, out of 3.2 lakh customers today, we have about 18,000 customers who have not paid us anything. So 18 divided by 320, that is about 5.6% in Microfinance.

In terms of loan book, how much would that be?

That's what I'm saying. So in terms of -- well, in terms of loan book...

See, that is lower.

It will be slightly lower than that, so about 4% in terms of loan book.

Okay. In Microfinance, it's 4% of loan book, which has not paid a single [ EMI ].

Yes. The good part is that restructuring should win some fruit. So out of the customers that we manage restructuring that had not paid us a single dime since this COVID thing started, about 27% paid their first installments in January. So I mean 27% to 30% is what we -- I mean, that's for -- as far as I'm concerned, it's nothing to write home about. Of course, I would have preferred the number to be much higher. But remember, these are customers who did not pay you a dime for the better part of 9 months. So getting them started is actually a good sign. So restructuring was in some way effective. The overall estimate is that this month, we might manage closing at right around 40% of the customer -- of the restructured loans. We get the repayment rate of the restructured loans.

Okay. Okay. And what are the total write-offs in this entire financial year?

Amit, the total write-off, both the companies -- I mean, on a consol level has been INR 7.04 crores, of which about 80%, about INR 5,30,00,000 or INR 5,80,00,000 would have been in Microfinance.

Okay. So I think in this quarter, it is around INR 3 crores, is it?

Yes. This quarter, it is INR 3 crores in micro.

Okay. Okay. Okay. And sir, what is the estimate? I think because now, I think it's been some time. So now what do you expect the COVID impact in terms of credit costs on the overall pre-COVID book that we have?

So it will be in the -- it will definitely be in the single digits. But that being said, I don't want to give a percentage, Amit, but probably in the higher single digits.

Okay. Have you also -- you've provided for about 7.2% of the AUM you've already provided and 1% is write-off, right? So already, 8.2% is what you have already...

Yes. But that will be some...

You have to understand, not all of that 7% will be available to me for COVID assets, right? There will be some that I'll have to keep on non-COVID assets as well. For example, there will be some on standard assets. So not all of it will be available to me when the time is there to write-off against COVID-related disruption assets, right?

So there could be further provisioning in the next few quarters?

No. So for -- as far as Arman and the Two-Wheeler and MSME, we have more than sufficient provisions. I don't think we are expecting to take any more provisions that are in place. For Micro, maybe one more quarter will be required.

And the net NPA number that you are quoting, these are -- if we include other provisions, then the net NPA number should be close to 0, right? I mean you reported about...

Yes. So that was actually my question, too, when I saw it. The thing is when you switch to ECL, at every bucket, you have to have a separate provisioning. So when you talk about 90-plus bucket -- overall provisioning might be enough, but the 90-plus bucket provisioning will be that 0.6%. Vivek, you have a better way of explaining it.

Amit, in terms of the MSME and Two-Wheeler book, the provisioning would be close to about 70%. So the net NPA would be the balance. 30% would translate into about 1.4% or something. And in terms of the Microfinance book, in turn would be about a single -- I mean, low single-digit in terms of percentage, the net NPA. That is without the Supreme Court dispensation being available. On the Supreme Court dispensation, obviously, I'm sure, like everybody, I think we would be 100% covered on the gross and net -- I mean, the net NPA would be 0.

So just to put it another way, prior to NBF ECL, the only provisions that were being created was on 90-plus overdue assets. Under ECL, you also have to provide, of course, a lot lower provisioning, but you also have to provide for standard assets, you also have to provide for 1 to 30 and 30 to 60. So counting all of those provisions, of course, you'll be covered in the entire NPA. But still, if you just consider the provision against 90-plus assets, you'll get that net NPA figure.

So -- and if you look at Microfinance in January, despite restructuring, we are at 92% collection efficiency. So how much of this impact that -- we still have this huge gap, is because of the Maharashtra book? And how much would be non-Maharashtra impact? So how much will be the collection efficiency in Maharashtra, for example?

Yes. So I'll give you December numbers. We have Gujarat at 95%, MP at 90%, Maharashtra at 85%, Rajasthan at 96% and UP at about 93%.

And these collection efficiency numbers are including arrears or this is billing efficiency?

No. I mean this would be -- so you mean like overdue amounts?

Yes. Yes.

Yes. Yes, of course. I mean this would include over -- you'll have to count it somewhere, right? Either it is previous months or in this month. So those would be included, yes.

But some prepayments and all that will not be included?

So Amit, the prepayment will not be included. But what happens is the simple way that we use is that, let's say, there is a prepayment of INR 10 on an original demand of INR 100. So the original demand is firmed up by 10 to the denominator and numerator, both are neutralized.

But the repayment rate -- actually, this is a ratio which I don't particularly care for because there are multiple ways to calculate it. And see, according to your question, so if you are counting arrears, for example. So if I don't count it somewhere, where do I -- either I'll have to change my repayment rate of previous months or I'll have to count it in this month, right? So honestly, you just have to kind of agree on one methodology and stick to it to make it comparable from month to month.

So I'm just confirming again. So INR 100 was due this month, and you got INR 92 -- no, you got INR 90 for this month and INR 2 from previous months, you are calculating the collection efficiency as 92%?

Yes. Yes. So you're basically using INR 92 as the collection amount divided by INR 102.

So for active clients, actually, let's say, customer has an installment of INR 100, and they are overdue 2 installments. So overall, in their current month, there would be INR 100 due for this month's installment and another INR 200 due for overdue installments, right, in this example. So for me to -- I would have to collect more than INR 100 to impact as what you are saying. So that will typically not happen, like the customer is not going to come and give me INR 300, right? If they pay me INR 100 against what is due at INR 100, then obviously, it is recorded against this month's demand. For me to actually collect on the arrears, it would have to be collecting more than INR 100 from that particular customer.

[Operator Instructions] The next question is from the line of Amit Mantri from 2Point2 Capital.

So just a couple of questions. One is your liabilities. NBFC as a promotion has declined from 29 to 21. Now we've been attending lot of the calls of these NBFCs that you borrow from, you've put pictures of some of the NBFCs. So they seem to be now refocusing on this segment or completely stop lending to other NBFCs. So what are you seeing there incrementally? Are these NBFCs lending to you, or you have to now look at other sources of funding?

See, to be very honest with you, one saving grace for us is that liquidity has not been an issue. In fact, we certainly get far more than what we can actually take. So right now, we are in a situation where we get to pick and choose. So obviously, we will go for the lower kind of -- lower IRR kind of low cost, which are coming from banks and they are coming from DFIs. That's not to say offers from NBFCs are not coming. They are coming, but for now there seems to be other better alternatives or cheaper alternatives, rather, I should say.

So then how has the cost of funding moved over the last 3 quarters? What is it now incrementally?

Our cost of borrowing is not higher. It has come down.

What is it exactly?

Amit, on a quarterly basis, it will be very difficult to calculate. But what has been seen is that a lot of resets have happened for the [ NCLRs ] with the PSU banks. And on a large number of loans, which were existing, it has come down by anywhere between 75 bps to 125 bps. On new loans, it is anywhere at least 100 bps lower, plus, as Jayendrabhai just pointed out, we've been able to get a lot of low-cost funds, in the sense, the DFIs like NABARD or SIDBI and MUDRA.

So overall, I would say our cost of borrowing would be approximately around 13.25, 13, somewhere around that neighborhood right now.

Okay. And on the securitization and assignment, that also -- it seems that banks are now reluctant to, especially in MFI, to do that. Are you seeing that?

We're seeing a lot of active interest, Amit. We are, at every point of time -- we have, as Aalok poined out, more active interest than you would probably be willing to take annual drawdown right now. Hence, we have a more situation of where we can pick and choose from the lowest cost of funds maybe.

And any updates on the equity raise?

No. No updates worth reporting, no.

And also now on the growth front, so Microfinance, you have now slowly started giving to new customers, which -- and largely until now, it has been to existing borrowers who have completed repayment. But MSME and Two-Wheeler, you're still sounding quite -- that you want to go slow there. While when we look at the collection efficiency numbers, MSME and Two-Wheelers is where the numbers are quite good versus Microfinance.

So we are planning on going a lot faster now. The issue was a lot of the credit bureau data that we were getting was -- so for me to evaluate a customer, I mean, one of -- especially for Two-Wheelers, let's say, a lot of my primary reliance is going on the credit bureau data, right? So you'll -- I mean, what do you do in that case? Maybe you up the down payments, you up the underwriting and those kinds of things. The same thing was there in MSME. So a person could come with a completely clean credit bureau report, but you don't know whether he has defaulted, actually defaulted at some other company or not. So it just required sort of a more deep dive into the numbers of the customers. But now we are a lot more comfortable. Of course, that still credit bureau data that I was -- that Jayendrabhai was also mentioning, that is getting back online and companies are, of course, reporting, maybe not as much as enthusiastically as they were pre-COVID. Of course, people are still restructuring and things like that. But at least the data available from the credit bureau side is a lot better, right?

So are you...

Yes. Sorry, go ahead.

So are you now starting to lend aggressively in MSME?

I mean I wouldn't call it aggressively, but the numbers are increasing. So our pre-COVID max was about INR 17 crores. Last month, it was about INR 14 crores. This month, we'll probably reach about INR 15 crores to INR 16 crores. So by this month or next month, we should be back on pre-COVID levels.

And can you say a number of what percentage of your MSME customers are already borrowing from someone else, from either MFI, bank or anyone else?

I mean I would really have to pull that out. I don't know it on top of my head. Typically speaking, in the MSME space, almost 30-odd percent would be debts. So 70% would have other loans.

Safely -- can safely assume that 30% are first-time borrowers, 70% surely have borrowed or have outstanding to somebody else.

Yes.

And what percentage would be, say, your existing borrowers, whom you have now shifted to the MSME book?

Those are few. I mean less than 5%.

Okay. Okay. And look, I think maybe a year back, one of the things that was planned from the growth side was the [ LAP ] or a rural [ LAP ] that you had mentioned. So -- and what's the -- any plans going forward on that?

No. I mean I think if you recall, we had shelved back for a bit post different kinds of prices. And of course, we have not taken it off our shelves post COVID. But this is definitely something that we can explore starting Q1. We have been exploring other individual kind of loans in the MFI segment. So these would be the customers that were with you for the past 3, 4 years and kind of similar to the MSME, but of course, the ticket size would be slightly lower.

Okay. And now after both Arman and the industry have taken this COVID hit, and this has come not just a few years after the demonetization hit, what does it mean for the Microfinance business? Does the business model have to change in any fundamental way? Or things continue to be the same?

I mean I don't know [Foreign Language]. I mean that's a very practical and a philosophical question. To a certain extent, after -- there is no -- no good crisis gets wasted in terms of microfinance. So there's always something to learn and improve. I think post demon also, a lot of people shifted towards more cashless disbursements and things of that sort. I think the next big focus for the MFIs will be cashless collections. But that is a challenge which is exponentially harder than cashless disbursements. So already, we're putting forth efforts in doing that, switching over to UPI-based payments. The response right now is very limited from the customers, honestly. But we have tried to tie up with many of the payment banks as well, such as Fino Payment Bank (sic) [ Fino Payments Bank ] where the customers can go and deposit the money at their local merchants or our FOs themselves can do it. So I mean overall, all said and done, even post this crisis, it seems that our customers are definitely resilient. I mean nobody enjoys seeing numbers like 92%, 93% repayments. So that means that 90% of my people are back to track and they are repaying. And remember, you are dealing with the most vulnerable people. If you and I, we lose our jobs, we can probably pay our EMIs for 3, 4 months out of our savings. This is not the case with our customers. If their income stops, then the EMIs stop. So overall, I guess, my confidence is somewhat boosted that this is -- I mean, we call demon as the worst-case scenario. This is at the expense of using the same terminology, exponentially worse than what demon ever was, right? And even then, if we can get away with maybe a year's worth of profit, I would say this is a resilient sector.

But so even now when we talk to a lot of the other players, larger industry players, be it, say, CreditAccess, Ujjivan or any of the other guys, they are still talking about aggregate credit costs of 5% maximum and not more than that, which just puts it in the demon category and not exponentially higher at least. So what is the disconnect? While you are saying that this is much, much bigger and the industry -- the larger players are saying that similar kind of credit costs as demon.

See the -- I think if you ask for people's opinion, you are going to get just that, different opinions. I'm simply giving my opinion. Ujjivan is -- none of us is right or wrong. But to say if this event is -- the magnitude of this event is in line with demon is a little bit out there for me personally. But this 4%, 4.5%, hit on a static pool basis during demon. So I don't -- of course, I really don't think that our losses are going to be limited to 4%, 4.5% on a static pool basis. Now by the time all the provisions are in place, if we grow to a level where the denominator increases, sure, maybe there might be 4.5%, there might be less, there might be more, who knows. Anyway, I mean -- so [ collect ] will continue for [Audio Gap] already, we have an [ RO ] structure similar to what we did with demon. We're still attempting to recover even 18 to 24 months after the event. So it's not like even if we write it off, we are going to stop attempting to collect money, right? That's not going to happen in the next [Audio Gap] for a long time to come.

Okay. And no, and you're saying that even for those 90-plus overdues, many people are coming back after restructuring, so -- which probably was not -- you were not seeing that kind of recoveries post demon, maybe. So it seems that there might be a longer repayment cycle for many borrowers.

In fact, the entire restructuring philosophy is behind that model only that given the kind of pandemic situation and wherein it has almost extended to 9 months now, and if you give them a breather, there is definitely a probability that they might want to fall in line over a period of time because microfinance funding has been their perpetual source of funding. So they would definitely want to fall in line once their incomes kind of replenish over a period of time. So the hard part is in micro is, of course, losing that discipline, right? So you constantly have to be in touch with them. But some of these -- at least 1.5% to 2% of our overdue -- of our customers are nowhere to be found, as in their house is locked, and they're, what we call, absconding in our terms. Those will be very difficult to recover, I mean, where you can't even find the customer. They are migrating someplace else or I don't know what exactly the situation is. But -- so those will be among the 4-odd percent -- 4%, 4.5%, which have not started yet because we can't even find them.

Okay. Okay. Okay. And good luck. Hopefully, from next quarter onwards, things will become normal.

Yes, I think so. I think the overall -- my overall goal is that from April 1, everything -- I mean, you hit the reset button and everything starts fresh again and, hopefully, forget that COVID ever happened and really kind of move on from a business perspective.

So just last question. What is the growth you're looking at next year?

Ask me again next -- in next quarter call. I am not sure yet. We are working on it. We are working on it. But there are a lot of challenges.

Amit, on a statistical basis, what would happen is that 2021 is going to be a slow year in terms of disbursement. So again, though we want to show growth based on what happens in 2021 or should it be almost like a full blown exercise to see what can be achieved in 2022, it's something maybe another 45 days away once we kind of move on from March, things would [ move ]...

And maybe around 8 to 10 branches will probably never start disbursements again. Those will be the worst hit areas. So at least the -- maybe it might not start again or we might not start it for a long time. So that will [ repull ] us to replace those branches with other disbursing branches. Of course, I'll have to keep the older one manned, so that will increase operating costs and stuff like that. A lot of consideration there, a lot of things to consider. But we are actively working on it.

The next question is from the line of [ Siddharth Bandari ] from [ I-Vest Global ].

Sorry, my questions have all been already asked. So I have nothing to ask. Thank you.

The next question is from the line of Debashish Neogi from Digitian.

See, COVID is a once in a 100-year kind of black swan event, and you couldn't have done anything about it. And now you're putting all the nuts and bolts so that you provide adequately for this year. And going by the collection trend, it's improving slowly but steadily, which is a good sign. And you're putting all the pivots for next year's growth. Now my question is from next year's perspective, in terms of growth -- see, the underpenetrated MFI regions, whatever I know, is Maharashtra, Uttar Pradesh, Madhya Pradesh. In these places, what is your plan for opening up branches for next year? And what is the reason of opening up in Haryana?

So see, Haryana was very specific. Specifically, we were thinking about opening in certain areas of Haryana, especially around the UP border. So western UP for us is performing amazingly. So right now, we have about 16- or 17-odd branches in western UP. And I think about 11 or 12 of them have already reached 100% kind of a repayment rate. So that has been performing quite well. I have a really good team down there. Also a few people who are highly experienced from Haryana. So it was a good time to expand into Haryana. And it seems that a lot of the other MFIs in the areas that we have opened or are planning to open, have fared quite well during COVID as well, as far as the repayments are concerned. So the other places that we are opening, it is in Rajasthan. So Rajasthan has also performed very well post COVID. The repayment rates for us have been 96-odd percent for quite some time now. And this was basically the -- Rajasthan was actually one of the first states for us to recover on a 90-plus kind of repayment rate post COVID. So these are the 2 areas that we are thinking now. Rajasthan has a lot of scope because we just moved in there last year. So there are a lot of branches that we can open there. Haryana, I believe, will be limited to probably a max of 10 to 15 branches. Other areas are in -- under consideration. So there are definitely a lot of areas in Bihar, especially in the Northwestern side, which have performed quite well. Other areas might not have performed too well. So I'm not sure. Gujarat, we are probably -- every place that we have wanted to open branches, we have opened. Maharashtra, I don't know. I highly doubt that I'm going to open any more branches in Maharashtra for at least another year or so. In MP, we can consider. MP [Foreign Language], the primary area that has been impacted is the Jabalpur region. Rest of MP has performed quite okay. So other areas we can consider, just try to -- we're trying to stay away from Jabalpur or any of the impacted areas. Basically, I mean, COVID, you're right, it's a black swan event. Nowadays, I have seen so many black swans that all I see is white swans. But the fact remains that -- it also gives you a lot of good data. My prediction is that the customers that have continuously repaid even during the COVID disruptions, I mean, that will lend a lot of confidence to the other lenders, especially people like myself also, right? So it might be feasible to give a higher ticket size to those customers or move them towards more of an individual loan, which is what our thought process is for the individual loan side in the Microfinance division.

And Aalok, this is a beautiful business to be in. And COVID has actually distorted all the figures. But before COVID, Arman has been on all parameters, has been doing exceptionally well for the last 3 to 5 years. You don't get a business where PAT grows at more than 40%; AUM grows at more than 40%; you have an ROA of 5%, 6%; ROE of 25%, 30% consistently. Now my other question is, earlier, the black swan events, whether it's demon, loan waiver, IL&FS crisis, used to happen. These black swan -- the frequency of event has off late reduced. Earlier, it used to be once in maybe 4, 5 years. Now it looks like once in every 2 years. So how you prepare yourself for it? Doesn't it make sense you overprovide for such events? I assume that ROE is not 30%, maybe 25% and provide for it a little more than normal?

See, there are definitely policies that you can frame to kind of expect regional level black swan events, which we do, right? So in any district, we don't have more than a 5% portfolio. And in any one branch, if it reaches INR 5 crores, then we split that branch. So from a risk framework perspective, we definitely take steps to reduce risk from a geographical perspective as much as possible. So if there are certain fires that erupt in certain regions, hopefully, that one specific event is not going to impact us too much. But when you talk about a national level event like demon or a global level event like the present COVID, I really don't see what I could have done different in FY '19 or something to really prepare myself, right? I mean short of taking pandemic insurance, I'm not exactly sure what we could have done differently. But yes, as far as provisioning is concerned, I think you are right that it's probably safe to build up provisioning, let's say, from 1% to 2% or something like that. In the MSME, we have already started taking more provision than we needed even before COVID because there was adequate profits and everything was in place, and maybe it was considered to be a slightly riskier business, although I don't see how given our performance. So yes, definitely, once we return back to profitability, that is one thing that we can consider as a higher provisioning. But again, the event like this, when you might have to write down 7%, 8%, I mean, how much -- I cannot keep a 7%, 8% provisioning on my book, right? That's impossible. I mean I would have to...

Of course. COVID is an exception. I'm not talking about COVID. I'm talking about more frequent kind of black swan event, apart from the COVID, whatever has happened in the last 4, 5 years.

So I said this pre-COVID also, and I'm saying it now that even on a pre-COVID, the time of microfinance with 100% repayment rate has gone. It's no longer there. It was there because everybody was targeting lower-hanging fruits and customers had a different kind of a culture where they were not used to getting easily available credit. Today, the culture has changed, and credit is readily available, and there is a lot of competition also. So when you add all of those things into the mix, it's definitely -- even if COVID didn't happen, you could not reasonably expect a 0 loan loss kind of a figure. It just wasn't going to happen. So in any case, we were ready to accept maybe 1%, 1.5% loan loss on the micro book. Going forward, we can probably provide for 2% or something like that on a steady basis. I don't know if that...