HDFC Bank Limited (HDFCBANK) Earnings Call Transcript & Summary

Hello, everyone. Welcome back to the Morgan Stanley Virtual Financials [ trip ]. This is Sumeet Kariwala, India Financial Analyst, and I'm glad to host the next session with HDFC Bank. From the bank, we have Mr. Srinivasan Vaidyanathan, who has been the CFO of the bank for past 2.5 years. Prior to this, he had a long career with Citigroup for almost 27 years, and he has held various regional and global leadership roles across diverse geographies, such as Singapore, Hong Kong, New York. He's also joined by Ajit Shetty from the Investor Relations team. Before I hand it over to Srini for some opening remarks, I'll read out the company disclaimer, a couple of them. First, there are participants in the session who are not amongst any research department, and their views can differ from that of the department. Also, this event is for Morgan Stanley institutional clients and financial advisers only. This is not for members of the press. If you are a member of the press, please disconnect and reach out separately.

So with that, let me turn it over to you, Srini, if you can start with your sense of the macro. How much has that impacted banks so far? What will make you worry? And apart from that, if you can touch upon the RBI lifting of the ban. And what can we expect from HDFC Bank now, what changes?

Okay. Okay. Thank you. Thank you, Sumeet, and thanks to the participants for coming in, and we're grateful for the opportunity to talk to you all. Okay. A couple of things you touched upon. Let's go first on the macro, what's going on geopolitical and what it means and what's impacting and so on, right? Indeed, it is something that is getting closely watched. But from an impact point of view, at this moment, while it is minimal, it remains to be seen how long it persists, and so thereby, we can see what it means, right? But assuming that in less than a quarter's time from now that this gets resolved -- the geopolitical issues get resolved, if you think about it like that, then what it means is like this. So you can look across 2, 3 dimensions. Look at it from a total overall growth dimension, right? What does it mean to the country's growth, GDP, et cetera when we are at this kind of a -- it's mostly external influence, right, in terms of oil price. That is the biggest kind of an influencer and some kind of trade restrictions that happen, right, along -- coming along to that. The GDP can -- our economist view is that it can be impacted by, call it, from current estimate of 8%, 8.2%, can be 7.5%, 7.7%. Call it, 0.5, 0.7 percentage points impact can happen. And where does this impact come from, right? If you see -- if you take that forward to see how do you link it -- link this up? See, the personal consumption component of the GDP is, call it, slightly under 2/3, about 60-odd percent, right, personal consumption. And about slightly 120% is the component of the fuel and transport and things like that, component in the personal consumption. So that is where the impact is, right, from a cost-per-share point of view, it comes there. And so thereby, the disposable income goes down, net of this incremental cost that comes from here. And so the other spending that happens from the personal point of view, that's restricted. So that's where you will see this impact of either 0.5 to 0.7 percentage points on the GDP that can be impacted from there, right? And then the other contributory factor would also be the global trade, right, export, import. Even if that gets impacted at an aggregate level by 1 percentage point or so, that's a net -- the net impact on the GDP that you see, right, impact. But still, with all of that, we -- our economist house view is that we expect 10%, 11% or so of credit growth in the following year coming. Then that takes you to the next aspect of the macro impact, right, okay? This is, what, from a GDP point of view, call it, color, what is going on with the inflation and rates and so on and so forth, right? The inflation in the short run is expected to be spurting beyond the 6% mark, right? It could be -- our view is that it could be as high as 6.2% or something is what they are estimating, but it's still a moving number, right? We don't know where it will end, but that's the kind of north of 6% is what we see. So the first half of the year could be somewhere around that range, right, given the assumption that these things persist for another quarter or so, that's the kind of range. The second half is expected to come back to around 5%, 5.2% kind of range on the inflation, the second half of the following '22, '23 financial year. So we expect, in the short term, inflation to go up and then come back down in the second half. From a rates point of view, we believe the RBI has always been accommodative, encouraging credit growth in the system. And just as things are happening before the geopolitical issues, just the things are happening, that I don't think that there will be, in the short term, any kind of a rate kind of an action; one cannot rule out. But that June is kind of a time frame we believe that one has to see what will happen to-date, rate scenario, right? Maybe that's the kind of time frame we should see. We will know more in a month or 2 as things progress and how this unfolds and settles. So that's from a macro point of view. Then the third-order effect in terms of what it means to credit and so on and so forth, it's too early, right? A quarter of these kind of things don't necessarily make a big deal of an impact. These are all transient things that can come back to robustness. But we'll have to wait and see. It's too early to predict anything on that front, right? We're closely watching that front. From a price point of view, the early indications that we are having is that to the little extent that certain things are percolating into the overall cost of inputs, the businesses, particularly the small businesses, are able to price them and pass it off, right? At the end of the day, it is about our -- ability of our customers. I've talked about the corporate customers or the CRB type of customers right now. Whatever the input costs are up to the extent that they are able to pass on means that hardly there is an impact, then it only gets down to the ultimate consumption type of impact there. So that's where -- that's the early stage at which we see right now. But we'd love to wait and see a little later how this turns out. Then the second aspect of what you asked is about the digital and what it means and so on. See, on the digital front, yes, we did -- after a few quarters, we got that lifted. There are a few things that we've been working, and there -- which rightfully we've been both addressing it from a customer point of view in the digital factory and in the back-end systems point of view in the enterprise sector, which we have been working on. They are progressing well. Some of them are in various stages. For example, the customer experience, which is coming in phases, came up in January, February, the Phase 1, Phase 2, which is a small kind of an implementation, not yet seen in the overall market. But in the -- in a quarter or so, that should come. Similarly, we are progressing well on the payments hub that includes the revamping of the PayZapp, the SmartBuy and the entire payment system as such, right, to embed into one and make that into a robust platform. That's progressing well, maybe a quarter out that more -- got out more that we need to be patient before we get to a close user group and then get past that stage and get to a public rollout of that to replace the existing PayZapp and come up with much more better functionality. So like that, there are a few things that, at some point in time on the call, I'm sure you touched upon. We can talk about the various technology initiatives, but at least, it's a step in the good direction. And it has opened up -- as we have always said, there is always some room for learning and room for bringing in robustness, which is what we focused, and we brought a lot of resiliency. And we strengthened our capacity. We strengthened our processes, or particularly, the processes that needed to be strengthened, we did all of those things. And simultaneously, we developed applications, which, imminently, there are some rollouts coming. And so we're quite confident of taking this forward.

Srini, I've got -- I've started getting questions from investors. What I'm going to do is that I'm going to punch them up and start with questions linked to revenues first because that's the key focus area. So let me start with loan growth. And what you've seen over the past 2, 3 quarters is the retail loan growth has continued to accelerate. If you can spend some time within retail, which segments have normalized? Which segments are yet to recover? And in particular, if you can talk about vehicle financing and unsecured?

Okay. See, from a loan growth point of view, you have to look at it in 3 segments, and then you touched up on the vehicle segment too, but that is part of the, call it, retail segment. But yes, let's touch upon that also. So you have to look at it in 3: retail, commercial and rural and the wholesale corporate. On the retail front, you've seen that quite a robust pickup is more of our choice, right, in terms of credit policy, opening up and getting back to pre-COVID level to get on to this, which, for most part, we are there. The cards was the last one that we were opening up and reviewing for the credit lines and making things available, which between December, January, we did. That's a cycle. We don't do it one big bang approach to it. So we phased that in over time. But other than that, from a loan origination point of view -- disbursals point of view, we're back on track in terms of what we were doing. So you saw last quarter a sequential momentum of 4%, 4.5% or so. We're building further on that, right? So that's where we are going. And then within that, you touched upon the vehicle segment. Vehicle segment has been hampered through various supply chain disruptions in the past and also in the recent time period, right, even after these geopolitical issues that are happening in the Europe area. It is even more an issue from that sense. So it's more a supply side constraint. There is adequate demand that is here, but there is a supply chain constraint, which means there's a lot of waiting time for delivery and so on. But from our growth point of view, that has been subdued last quarter. I think we reported 1% or 2% growth in vehicle when the industry was shrinking, right? The industry had a degrowth, but we had a 1% to 2%. And similar kind of is what we are seeing even as we see now, right? The industry is having negative, and we are trying to stay afloat and be there at the margin in a low single-digit type of thing. That's where we are focused on, and we continue to sustain at that kind of level. But other than that, if you look at the rest of the retail product, it's broad-based, right, across secured, unsecured, which is mortgage, Gold Loans or unsecured personal loans. You're seeing across all of them that we saw last quarter. We continue to progress -- make further progress on that. In the commercial and rural, that was something that is important to the country and the big power for the GDP growth that you are seeing, which current quarter, we expect the -- current year, we expect the GDP growth to be closer to that 8% change, right? That's the kind of a rate of growth that we expect on the economic growth point of view. And the commercial and rural is an important kind of a driver for the country. And that -- you saw last quarter, we had 6%, 7% sequential growth, and we're confident of maintaining that kind of a momentum in terms of growth across various products that's there. So that's going fine, both part of geographic expansion that we are doing, both in terms of districts in which we operate and we offer. And the villages where we offer these products, that has expanded and, of course, supported by additional relationship managers because it's not only getting for a broader wallet share from the various supply chain segments, it is also getting new-to-bank customers. So we are bringing in several new RMs in the area to make expansions, to bring in new customers, too. So that is doing fine and sustaining that momentum that we have seen. On the corporate segment, you saw last quarter about 4% or so sequential growth, which, historically, if you see, historically, means not just the last 2 years of COVID. Even the year before COVID also, if you see, we were growing in high 20s, 30s type of percent. It moderated to a sequential momentum of 4%, call that, it's a mid-teens annualized, if you do. That's a rate at which -- and we are really okay with that because in the past, we were risk-off on the retail and commercial, and the highly secured, the highly rated corporates, we were quite happy to do. And now the rest have taken off in a good kind of robustness for growth, and this moderated to 4%, which is still very good but moderated. And we are happy for that mix change to reverse itself where we get a higher proportion of retail and commercial. We are happy with that mix change to come, which is what we are -- we saw last quarter, and that similar trend is what continues.

Srini, let me ask a couple of quick follow-up questions with respect to loan growth, and then I'll get to the fees. I can see a lot of questions around that. So within loan growth, are there segments where your risk appetite has not normalized even now?

No, I wouldn't say anything that risk appetite is not normalized. I wouldn't say any segment. Cards was the last one. As you would imagine that you would think that the customer, particularly with the retail segment coming out of such kind of a COVID type of crisis. First, COVID was -- first wave was economic, and then the second wave was both health-driven and economic type of crises. You would expect that the customer when they come back, first thing is they would like to go into secured product because that's the least cost of borrowing. Then they will go into an unsecured, which could be in the low double-digit type of cost of borrowing. Then they'll come to a card type of borrowing from a customer behavior point of view. And so we wanted to be careful and not allow customers who could borrow at a higher rate come right in the front and do because we wanted to be careful in opening it up and which we opened up right now. So I would say that across all the segments, we are operating at the pre-COVID type of levels.

Okay. So let me try and tackle this from the other side. From a demand perspective, in the unsecured personal loan segment and credit cards, where are we in terms of demand? Have you seen that come back quite well? Because now we've normalized as an economy, so is that back?

Okay. Unsecured personal loan, we saw last quarter itself they're good to come back in terms of demand and that continued to sustain some growth from there, right? So we have seen that come. On the cards front, it's a cycle. What happens is that first is you get the sales done. So the sales grew by 24%, 25% last quarter, you saw. But the sales are more driven from customers who are transactors. So you saw the loan growth, only 10%, right; sales growth, 25%; loan growth, 10%. So people -- more transactor types who have utilized the balance sheet and you don't earn much income and then 10% loan growth, right? That -- even the 10% loan growth, it doesn't necessarily mean they are interest-paying 10% growth. It is total growth that's 10%, lower than the top line growth you are seeing, right? And as we see now, very only -- but I would say that the value that takes you to think about the card that you asked in terms of loans from cards, right, there can be a good amount of growth coming from cards on the balances. Then the interest-paying customers, right, how robust they are growing. That customer behavior is marginally improving, but it is nowhere close to pre-COVID type of levels because you would not expect -- I would not expect them because first is they will exhaust their low-cost secured. Then they will exhaust their kind of an average cost, call it, in low double-digit type of credit. Then they will come into a little more higher cost of credit in terms of what they want to borrow. I wouldn't say that their gut area is back to fully pre-COVID levels yet. It is coming there. And we have one more type to overcome, right, which we had a vacuum that we had 8 months of -- 9 months of vacuum where the cohorts were not building, right, because as the old customers go out, for various reasons, they go out, the new customers come in. The rate at which you bring in, in the normal circumstances, right, that some customers go out, the new customers come in, the rate at which the new customers come in always exceeds the rate at which people go out. And so then thereby, you have a good build going on. We didn't have 9 months of cohorts coming in. But we have accelerated that cohort buildup. We had 1.3 million cards, and we continue to have accelerated buildup of new customers coming in, having credit card. But it will take 2 to 4 quarters for those customers to mature, right? You activate them in a quarter, get them to spend in the following quarter. Then you start to increase the spend, and then once they increase the spend, then you get into having the kind of revolving or interest paying and so on and so forth. So that is the cycle that takes 2 to 4 quarters to happen. We have begun that, so we have to wait and work through that cohort buildup again on the cards. Yes, a long answer I gave, but the short answer is personal loan, we are seeing that. In the card loan, we don't see it back to pre-COVID level, but we are seeing good signs of coming back.

Srini, could I ask you to elaborate on that credit card explanation that you gave? And there's a question from one of the investors that if you look at the fee income decline that HDFC Bank saw, it's slightly somewhat higher than the other peer banks. And I think so you, in a way, alluded to one of the reasons around that, which was 9 months of credit card [ band ]. So can you talk in terms of what has been the decline in the rural rate for you, vis-à-vis the industry? And how much time will it take to come back? Because I think that could be one of the drivers, which is not present in the case of other banks.

Okay. Very good. Okay. See, our rural rates, if you see, I think we did allude to it last time, which is they are at like a 70%, 75% of what we have seen pre-COVID, right? Rural rates have come down that much. But what is -- whether it is in line with the industry or not, I won't know because there is no standardized industry reporting on revolving balances, right? So I won't know. But all I can tell you is that there is no peer benchmarks or anything that we can allude to. Our credit card book is about, call it, INR 70,000 crores, right? Definitely, more than INR 70,000 crores. And if you look at the peer kind of a credit card book, probably slightly under 1/3 of that size, right, less than 1/3. So I do not know what kind of a benchmark or comparisons we can do. We have not been able to allude to it. But then the size is very different, less than 1/3 or so. Similarly -- and when it comes to fees, you don't look at fees in isolation, right? It is interconnected to several things, particularly when you look at the card, the credit type of fees, right? So you get a customer who spends some things and has got some revolving. Then the same, similar kind of bunch of customers, some of them pay the amounts late. And the late-paying customers, of course, you want them to be paying on time, but some of them, you don't mind paying late because they are good customers. That is part of the credit model. That's part of the credit cards business model, right? Some of them, you would want them to pay late, and they don't mind paying the fee. So that's a part of customer base. But over a period of time with customers who are habitually late in paying, they have been stressed and -- or in the NPA bucket, or they have been in restructured bucket and so, either with us or in the bureau. So our bureau has told them to get them out and stop, right, freeze all the credit lines. And some of them will come and take cash advances and take the fees. Again, cash advance is a high-cost borrowing for a customer. And if that is the kind of a customer you would be exhaust as other things before coming, and if this customer is still doing, then our credit has tightened that aspect of allowing those credit lines. So much so -- and some good customers who we have that is part of couple of other statistics that we have put in the market to show that if you think about the credit utilization, credit line utilization are at a 60%, 70% of a pre-COVID level, credit line utilization, right? So that means the people are in to foster while the spend is growing at 24%. There's still a long way to go for that to come because things scale down. And then now when you look at the year-on-year, it is showing good 24% growth, but it's still long way to go back to pre-COVID level from utilizing the lines available point of view. So that is one thing. The second thing is the liquidity in the hands of the customers, right? We have a little more than 70%, 75% or 70% of customers who are our own customers who have a liability relationship also with us, right? The ratio of card loans to deposits is 5x, right? And pre-COVID, it was slightly under 4x. So the people with whom we do business seem to be sitting with enormous amount of liquidity and with us itself. Then in the banking system, they should be having much more liquidity with that. With [ that itself ], what was 4x now has become 5x. So people are sitting with good amount of liquidity. And so you -- that is why you see some of those kind of credit-related type of fees are down. And then last quarter, I also mentioned about there were certain other promotion types. So typically, promotion types go with some kind of offer, right? And so that's why the fees were down. And similar to what I had mentioned on the revolving, how long does it take and when does it come, we should wait because it is related to the credit of the customer, the credit behavior of the customer, right? And we need to be patient enough for 2 to 4 more quarters to say how that comes back up, and it is also the absence of a cohort also impacts us. We were a little more impacted due to absence of 8 months of cohort that was not there. Otherwise, there would have been some level of offset that would have come, but it wouldn't have still changed it in a significant way, but there would have been some kind of a plus that would have come that didn't come, and now that cohort is also building up.

Srini, let me move on to fee income with that discussion. So what we have seen over the last 2, 3 months -- last 2 months, actually, Jan and Feb, there's a reasonable slowdown with respect to sale of third -- insurance products, and that's obviously a very significant part of the income, particularly for the fourth quarter. Has that happened at the bank level as well? If you can throw some light on that.

See, typically, our, call it, the third-party product or the fee income, right, normally hovers between high teens to low 20s type of range in the insurance. That's the kind of -- right? And we are quite confident that it is operating at those kind of levels. Month to month, things can happen very differently, right? You alluded to last 2 months, Jan, Feb or something like that. We cannot look at 1 or 2 months to look at a quarter, and even one quarter to another quarter is very different. But normally, March month is typically a seasonal quarter for these type of products due to various -- I think it is connected with some tax rebates and so on and so forth, too. And March is quite normally very robust on this. So one has to wait for the quarter to be over in terms of how the customer is thinking and doing transactions. But I wouldn't let you just think about 1 or 2 months; you have to look at the total quarter. But from our portfolio point of view, that is the kind of insurance performance that we have done as high teens to low 20s type of what we have normally.

Got it, Srini. So March is important, and we should wait for that. Fair enough. Srini, there are 2, 3 questions with respect to fee income and how should we think about it because there are a lot of variables that you're talking about. You're talking about revolvers coming back in 3, 4 quarters. There were some fee income waivers, which you gave. We don't know whether that will repeat. You're talking about retail credit clawbacks, where rating that it drives some percent in fees. So overall, when we think about next 3, 4 quarters, fiscal '23, what kind of fee income growth should we expect?

See, while we -- normally, we don't say this is one particular outlook or something. We don't give an outlook as such. But what we have said is in terms of fees, right? Every time we come to a fee, people ask about what should be kind of a normal range of fees for us. We always say the normal range of fees is in the kind of mid- to high teens is what we would say normally. That should be the normal, long-term sustainable level of fees, right? That's how we think about it. And in a particular quarter, if you are in the 20s, there were certain other important things that came or kind of a customer propensity to certain products that came through. Otherwise, the way to think about, from a modeling point of view, if you ask me, is think about mid- to high teens is what is sustainable in the longer term. [Technical Difficulty]

Okay. Just sticking on to fee income, there's a question with respect to MDRs. We don't know whether the RBI will reduce that or not. But for example, if the RBI were to reduce MDR by 25 basis points, what kind of impact that could have on fees?

Okay. See, there are 2 things, Sumeet, to think about on the card's MDR, right? One, MDR as such, and the second one is the interchange, right? You can't look at only India. You look at things, let's just say, MDR and interchange. MDR is something on the acquiring business and interchange is in the issuing business. On the acquiring business MDR, we've always said we don't make money at all. It's 1 basis point or 2 plus or minus all the time. That's it, right, in any given quarter depending on the mix of how it goes, a few basis points, plus or minus. That's how it operates. So that means it is not a stand-alone profitable value proposition. It doesn't. We are in that business for a broader relationship with a customer that brings in the liability relationship, that brings in the asset relationship and has the engagement from an overall product and the ecosystem of the merchant relationship, right? That is what we are interested in and we do business. We are the largest merchant acquirer in the country for that reason, with a little more than 48% market share, right? And with the 48% market share, we don't make money on the MDR, we don't. Now that takes you to -- so if that changes, then the dynamics on various other pricing will change. That is the point I'm trying to say. And then say, I suppose that dynamic now changes and impacts the interchange, card issuing business interchange, right, that affects all the players in the country. And our card issuing interchange, our issuing sales market share, call it, about 24%, 25% or so, right, issuing sales market share. And Sumeet, when it impacts your interchange on the issuing side, then what does it mean, right? That's the next aspect corollary to -- there's nothing on MDR. When it comes to interchange, what does it do? Looking at an interchange also in isolation is not right. Why? Because when you have an interchange, think about your own -- Sumeet's card that is there with the bank, right? And when you spend, there is something that we earn as an interchange as an issuer of your card, right? There is certain other costs that go with that, too. So you have to look at the holistic aspect of the P&L that comes from the issuing business. What are the other type of costs that go with? There will be some sales and promotion, there will be some interchange, there will be some franchise costs and so on. So there are several elements of costs that go with it, right? And so when this changes, there are enough levers to change various other things, right? And at the end of the day, it has to make economic sense. So if one thing changes, there's going to be the second-order, third-order effect change that needs to happen to make the entire P&L work can be profitable.

Fair enough, Srini. I'll stick to revenue growth for some more time. There are a couple of questions I'm seeing. One is with respect to interior India, rural, semi-urban areas, where you've been -- you've accelerated your customers in these geographies quite meaningfully. There are very few banks that will match you. What we hear about these geographies is that there's a meaningful amount of demand slowdown, which some of the companies have been talking about. What have you seen? And where are we in that side?

Okay. See, semi-urban and rural is a critical geography and is one of those fast-growing and focused geography for us, right? And part of our CRB vertical is also adding more emphasis into that kind of a geography. What we see -- I'll tell you from a low credit or even from a deposit growth point of view. We are seeing more balanced growth across various geographies. So we are not seeing any softness or anything like that, right? It was impacted during the COVID. And when it got impacted, it got impact to the lag because in the initial part, the semi-urban and rural was oblivious, and then later on, it did impact, and then it came with a lag into semi-urban and rural. And then when it picked up, things did pick up everywhere, right? So we are not seeing. And if anything, anecdotal, what we hear from our on-the-ground salespeople and relationship managers is that, there is more money at the disposal and they are also viewing more to come in time, right, the next short-time period because see, at the end of the day, this agrarian, semi-urban and rural is all -- think about it as all commodities, right? And generally, the commodities prices have gone up, right? And if you think about certain grains to kind of -- or any kind of other products that you see, the prices have gone up. That has put more in the hands of the people in this [ and they are with the people ] in the past. And it is also expected that this coming harvest season comes through, it is also expected to yield more in the hands of them. So from our point of view, from our banking point of view -- bank point of view, we did not see that.

Interesting. Srini, I have a question with respect to mortgages. And when we look at your large private banks, they do a lot more on mortgages. I know HDFC Bank doesn't originate mortgages in its own book. It [ does for HDFC ]. But is there a way in which you can participate more in the upcoming housing boom? What are your thoughts on that?

It's an excellent question, which I should have covered it the first part that you talked about the retail itself. You're right. There's now -- the mortgages, as such, is in a very classic cycle. Beautiful place where after 2014, '15, onwards, where it has been, for various reasons, hampered from RERA regulations, to demonetization, to NBFC crisis, to GST cost of production increase. So various reasons, right, along the way. It had been hampered. Now it has come to a kind of a situation. In fact, after the first round of COVID, that was the first product to take off and go in a grand way because all the government, central or state, everybody started to give a lot of incentives in the form of reduction in stamp duties or waivers and things like that to make inventory clearance faster for developers. So at this time, we do see that there is a big amount of housing demand, which has not been fulfilled. The inventory is coming down. The developers have become stronger. So we see that cycle classically taking off beautifully, right? We are participating in that. You will see that in our brand system, where here too, it is for various reasons was not a great deal of focus. So we have ensured that we are having a balanced approach to all the products, and that includes mortgages as a critical product for our customers, right? We have a very low penetration in our base, 68 million customers. We have very, very low penetration. So we wanted to take it up in a big way. And so we are moving in that direction to get our people trained, to get our people to have the right kind of engagement. And our credit and analytics is supporting our relationship management -- managers to have the right conversation with the customers, both from a -- participating in the new mortgage as well as any balance transfers. So on both sides, we are focused to bring that book, and we like that book. And the time seems to be great for that.

No, I can't agree more. With the margins, Srini, and you talked about RBI rate hike, and it [ didn't ] happen in March quarter. You alluded that your expectation is more like June. So maybe even next quarter will not have a rate hike for bulk of it. So when would margins really pick up for HDFC Bank in that backdrop? And can that pressure margins downwards the delay in rate hike?

Okay. Yes. So there are 2 things -- 2, 3 things to think about the margins as such, right? One -- the first one, you think about it is the mix of the products. We have come down, broadly Basel classification, if you see about 55% retail has come to about 45% or so round numbers, Sumeet, and then the wholesale is a reverse. It has happened. And that is what has brought the margin [ stump ]. And when the margins come down, the cost also comes down because when the retail goes down, the cost of delivery and the cost of acquisition, the cost of delivery, that also goes up. So you saw the costs are also coming down, right? And then that is how you think about from a -- think about our margins on a risk-adjusted basis. On a risk-adjusted basis, our margins have been a little more stable than what the overall headline margins suggest, right, because you price for the risk all the time. That is how you think about it. And so if you straight go down to ROEA, for example, 3 years ago, ROEA was, give and take, COVID time period, 1.9 to 2.1. And just call this time period as post-COVID, right, although Omicron came and went, but around that time, similarly, at 2, 2.1 type of range. So you get -- and how do you get that? You get that by various mix change. When the mix changes into more secured, high kind of rated kind of a customer, you have a lower margin, lower cost to deliver -- acquire and deliver, lower credit, and you'll come back to kind of returns that are robust and good, right, which are at the historical trend, right, so that kind of -- you maintain that trend. And when the retail starts to pick up, you start to get more margin, you start to get more cost to support that, and then you get more credit that comes with it, and then that's exactly what you price for the risk, right? That's how -- think about the last quarter. The credit cost -- the pure credit costs were 60 basis points or so, right? And we look at the cost to income, was 37 basis point -- 37% also cost to income. So it is about how we -- which is the mix of the product that is there. And then -- so looking at margin in isolation, this is not -- I don't think is right because you have to look for what do you price. That's why I allude to -- you have to look at the risk-adjusted margin, and then along with that, you look at the cost, even after risk-adjusted margin, and then you look at how the cost is tagged to those because you know that in the -- when the mix changes more towards lower-margin, highly rated customers, the costs are pretty benign because you get the high ticket, costs are very low. And the cost to income is in the teens, right, if not in some product lines in single digits cost to income. And then you see credit costs for some highly rated customers, virtually, the expected credit losses are nothing, right? And so that is how you look at the margin in connection with the cost and in connection with the credit also. And then you look at the bottom line, right? What does it all turn south to? Return, at the end of the day, what hits the shareholders from a returns point of view? Where does it stack up? So that's -- so coming -- the last aspect of where -- then you ask how long you think it should take. See, the retail pickup is happening. That means the pace at which the mix change is happening, I would expect 3, 4 quarters for that mix change to come back. It has to -- when it get back -- when it gets back to, call it, 2019 level of mix, right, which we have modeled, when the mix goes back to 2019 level of mix, the margin changes by 20 basis points or so, right? It operates in the mid-range of 4 to 4.4 that we have operated. When you go back to 2019, it changes -- gets back to the mid-range of the historical.

Srini, you made an interesting point that some of the -- because your loan mix has changed to that extent [ that it has ] come up on an underlying basis. But what you've done is over the last 3, 4 quarters, you've looked to accelerate tech investments. So when we look at the cost growth, it's still quite elevated. Actually, it's running above revenues. So is it fair to say that the acceleration in tech investments that you had to do has already happened? And F '23, F '24, there could be some moderation because the starting point is in the base?

Okay. I'll come to the -- take all because there were 2 other aspects to mention about the margin itself, right? I mentioned about one. The other aspect of the margin was also the -- which we alluded to, not only the mix change, along with that, the kind of a card mix, right? Because you know the card yields are higher than the rest of the year. So that has to come back is one other point I wanted to make and where the cohorts have to build itself up and come. And the third aspect of the margin, which is a minor impact is when the rate change -- because you alluded to rate change and then thereby the margin, when the rate changes, it shouldn't typically affect the margins in a big way. But whether they -- it can affect in the short term to some extent, yes, because that's a lead and lag effect, right? So when the kind of a baseline rate changes, you get the asset pricing done first, and then you follow with the lag, the deposit pricing, and then it goes, right? So that's a legal lag effect that you can do and manage that. So that is -- I want to leave that thought on the margin, right? There are 2, 3 other levers that have to play out and come. And now -- then you asked about the cost, particularly focused on the technology type of cost, right? See, we have spent a lot on technology, the cost on technology. And the cost of technology, you can think about it in 3 aspects, right, what it means. One is on the front-end type of technology, right? The second type of cost is on the core banking system's robustness on technology, right? That's the second aspect. And the third aspect is on the automation and productivity, which is more internal related. One is the front end than the back-end core processes and then the automation and analytics that is there more at the back-end aspect of it, which outsiders wouldn't see that aspect coming through from the technology point of view. Various technologies are in process of going to, right, which I said. On the front end, the 3 pillars. On the customer acquisition side, the Adobe type of products. Some are in. Some are coming in, right? The journeys -- the digital journeys that we are redefining. From the platform and products delivery point of view, right? The payments hub or the new PayZapp and things like that are in the works, imminently going to come through. Then on the relationship management pillar side, the customer experience hub as one other example of a technology that is getting rolled out, some phases of that are in but more phases to come, right? So some of those technologies are in, and some are to come. And then if you think about the enterprise factory, which is the more core banking overall, that has started. And with a bit of tech partnership, with a bit of tech start-up, we are co-creating certain things. That's a 3-year journey. Again, several modules need to come every other quarter. There are milestones to deliver those. And so that will be ongoing over the next 3 years. That will come, right? The third aspect of it, which is I told you about the analytics or the back-end automation, that is a continuous process that we have done and we will continue to do. So if you're asking me whether we are done with the technology investments? No, no, we are not done. Have we spent something? Yes, we have spent something. We consistently spend something. Now we will continue to spend on technology. That is part of -- and whenever we spend technology, there is also a certain level of automation, certain level of productivity, certain level of pacing that happens to manage the overall business growth, right?

Srini, If I can simplify my question, if I look at tech spend as percent of revenues, I don't need the number. But for example, whatever it was 3 years now from this year, last year or next year, you will see a significant jump as percent of revenues because you're accelerating that investment. What I was trying to check is that delta has played out in fiscal '22 all the way to fiscal '23, if you can give some color on that. Or it's both...

No, it has been -- yes, it has been quite robust. It will -- the overall tech spend will go up, but that doesn't mean that the tech spend is in isolation to be looked at. If some tech spend is going, there has to be certain other cost that needs to come out because what is technology doing? Technology is replacing something else, right? If I don't spend on technology, I will be spending on people, right? I wanted to manage the process. Now I have a technology to manage the process. Now I'll have technology to acquire the customer in much more elegant manner that gives a better turnaround. So that gives me throughput and scale, better scale it provides, right? So you look at -- looking at technology cost in isolation is not completely something to look at, right? What sort of people dependency versus technology dependency for -- to look at that, it is important to see but not necessarily because it is about the total cost that you need to worry about. If technology costs will go up, some other costs is -- you don't need to go up at the same pace and some other costs.

Fair enough, Srini. I've got 2 questions on asset quality. One is, do you see any challenges with respect to restructured loan book that was done in the second COVID wave? So that's one. And -- okay. I'll do the second question together. I guess you'll give your answers. Isn't the bank sending conflicting message when it says that asset quality outlook is strong but continues to make contingency provisions by keeping P&L provisioning high?

Okay. Two aspects. There are 2 different things, but let's take. The first aspect you talked about is the restructuring and how should you think about it, right? See, the fact that there was a restructuring asked for by the customer under a program, which we liberally went ahead and implemented because we want to be seen as helping the customers and not putting the customers on to the bureau as a problem customer, there will be some level of stress. There has been some level of stress which you have seen, right? And our -- the way we look at it is what part of it is secured and what part of it is unsecured, which I think -- last time, I think I gave a data in terms of what -- slightly under 50% or something is secured and the rest, unsecured. And then secured, we feel very comfortable. Unsecured then, we further mine that data to see what part of them are salaried and what not. And within that, we look at who -- which of these customers have credit scores, which are quite robust and good and which are not, right? And so thereby, we focus our energies towards that. Whether the restructured book will have delinquency higher than the normal historical, they will be there. These people had some stress. That is why they took some restructuring, right? Whether this is going to change the overall credit profile of the bank, we don't think so, right? We don't. That is why we said we always feel comfortable with the overall book, right? That is how you should view it. And that's how we work with the customer to look at it on a total basis, right? And then your other aspect of the question that you said was, "Hey, how should you think about the contingency provision?" See, think about the contingency provision that we created. Go back to 2019. There is no sign of any COVID in 2019, right? We built the contingency provisions. So when there are opportunities and we want to build the balance sheet strong, enabling for the growth, right? At that time, we did build it because we wanted to -- when the economy was going down, we wanted to get out of the block first to start because without the COVID, the economy would have started to take off after touching the bottom and go, right? We wanted to be ahead of the block to grow to keep the balance sheet strong enough for any kind of -- if anything were to happen on credit, we got a good amount of reserves. Then COVID struck, right? So we started to build it even before that. So in other words, if you look at before September 2019, our contingency results were less than INR 1,000 crores. And as we speak now, it is more than INR 8,000 crores, INR 9,000 crores, contingency reserves, right? So along the way, we built, and if there is an opportunity, we will build to keep the balance sheet more resilient because it allows us our front end to originate and our credit to make several experiments and so that we can get the right kind of product mix to the customers that we want to. So that is how you think about reserving, as something that you build the reserves to keep the balance sheet strong. If there were to be some kind of credit decisions that you need to relook because you made some experiments, so be it, right? We learn, and we move. That's the reason we build reserves.

Srini, have you noticed a difference between you and peer banks with respect to the accounting of cash tax? Like, for example, when you reported your fee income last quarter, you said that part of the fee income was also because of levers that you provided in the Festive Treats. Some of the banks actually likely take that via the cost line. Have you noticed something like this? And are you aware of any accounting differential that you guys follow?

No. I think we are following generally accepted principles -- ideal principles. I don't think there is a difference to the best of my knowledge on anything, yes.

Just to be clear, very -- just to be clear the specific-related aspects that you would have offered, you would have deducted that from the fee income. Is that correct?

No, not the -- no, if the bank provided kind of a cashback -- cashback schemes come from various sources, right? One, the partner merchant can pay the cashback. We can do the cashback promotion. We can offer rewards cashback -- in lieu of cashback, we could offer rewards. So there are different accounting treatments that happen for all of that. The waiver is when I originate a card loan, when I originate something and I will charge you -- when you want INR 10,000 of loan, I'm going to charge you INR 1,000. And I say, okay, I will not charge you INR 1,000. I'll charge INR 100. That is a part of the processing fee. So if you look -- if you go to a branch and see, they'll say, "Hey, I'm going to have 50% discount on processing fees." Processes fee waived in the festive period, right? You will see things like that promotion -- various promotion happening.

Right there, Srini. Srini, we have 5 more minutes, but I've got 2 questions for you. So -- and they're somewhat related, so I'll read out both of them. There was a media report on HDFC Bank launching digital challenger bank. Can you elaborate on that? And then there's a second question. Can you -- which is related in a way, could you ask about their plans for Digital 2.0?

Okay. First, let's take that other one that you asked about the digital bank. See, we talked about this, Sumeet, in the past in terms of partnering with the tech start-up to develop a digital-only bank. So this digital-only bank has got a few things to look at, right? You know the traditional bank that we have, you acquire, you service, you have a relationship management, right? So you have, and then there are several technologies that go to support all of these things. And then you have the back end that does various operations and service and so on and so forth. This digital-only bank is supposed to be a bank targeting certain customer segments. Call them, for lack of anything else, segments that are new-age customers, right? For lack of anything, call them new-age customers who are quite comfortable using a mobile to do banking. And generally, they don't need or they don't want and they don't prefer relationship management. They want self-service, right? So that is what this digital bank is supposed to accomplish. It is supposed to accomplish that it gives them certain vanilla products, helps them to operate on their own on a mobile with -- there is no back-end operations. So that means it doesn't -- it's not a front-end bank. Digital banking is not a front end. It's a end-to-end bank with zero paper, zero touch, straight from the time when a customer opens an account, thus the transaction goes, but it is envisaged to plug into our credit model so that it is simply not only a savings account. It has also got other lending value propositions in it. So from an API, it plugs into our credit model so the algorithm that we can run can be the same algorithm they run in our traditional bank, right? So that's what this digital banking is under development. We will roll that out at an appropriate time, but it is supposed to be a bank that can be something that supplement our traditional bank, right? The brick-and-motor, the digital bank that we have, it can supplement that. That is what is envisaged. And at some point in time, there is an outside digital-only bank that gets licensed. We are an incumbent already in the play. That's all it makes us. And it's a scale play, right, where here you have 68 million customers but only 11 million, 12 million customers with whom you have active relationship management. We want to scale up to 20 million, 30 million, right? So that is why the VRM channel came into being, and then 2 years, it has been little new form. Now it is getting back on the bigger scale on the relationship management side, right? So we want to take that up because that's where the profit pool is. As I said, you acquire, you service, you have relationship management. Close to 2/3 of the value for the bank comes from relationship management -- active relationship management, right? That's where you make money. That's where you deepen the relationship. You have a breadth of product that you offer to the customer, you make money. Then we are seeing that is why this digital bank is supposed to bring unit economics. So that means you bring in customers who don't want relationship management. At the same time, you want there -- you want them, then you work on unit, so you bring in higher quantity of customers come in. The unit economics are low, but then that is where you offer on a digitized basis and get the right kind of profitability that comes through it, right? That's the kind of dynamics. But yes, we expect to launch that at some point in time in the next few quarters. We are at it. So that's part of the digital banking. Then the Digital 2.0 that you talked about -- see, Digital 2.0 is a very broad term, right? That is we've had certain digital implementations in the past, starting from 2014, '15, various things that we have done. Digital 2.0 is a very broad technology to say, the next generation, the next kind of application that comes in Digital 2.0, what I can say there. You can think about this digital bank as something that comes under the Digital 2.0. But then the other aspects of not necessarily Digital 2.0 but the new foray, right? If you can think about what I had alluded to on the first pillar, which is the customer acquisition pillar, the Adobe partnership that defines -- redefines the journey for a customer, right, for acquiring the customer, it redefines the journey. Then in the servicing, it's something doing the new mobile banking app. We want to revamp the mobile banking app. That's one, bringing the payments hub, which is the new PayZapp, the new SmartBuy and bring in all of that and bring in some redundancy to the net banking and mobile banking and put that in the payments hub, right, to enable customers to do things in a better manner and in partnership with a tech start-up to get the latest UI/UX kind of attraction to it. So that's something in this one. And then on the relationship management, the Sprinklr is one example. And similarly, on the commercial banking -- commercial and rural banking, small business banking, we have some CBX and some trade, et cetera, et cetera, apps that are getting rolled out in that relationship management area. So these are all certain things that I already mentioned in the beginning of this call of certain technologies that are coming. You can broadly think about them as the new-age digital technology rollout rather than very high-level Digital 2.0, yes.

Actually, by when can we see all of these products that you spoke about? Is it a 1-year, 2-year journey?

Several of them starting in a quarter's time, right? For example, Adobe -- out of 80-plus journeys that we have defined, 8 or 9 journeys or something is already live, right? And the rest, one by one, will come over a period. We are not waiting for any big bang. It is day by day, it is coming. And then if you think about the payment hub, the new PayZapp and the SmartBuy and all of those that goes into it, that is an activity perhaps a quarter away, right, that will come. The new mobile banking app, maybe a couple of quarters away that will come because we want to get to the next generation of the mobile banking app. Sprinklr, as it only in phases, it is coming. Some phases have come. But the bigger phase is where the customer experience actually takes a greater shape, maybe a quarter to 2 quarters out from there. So various things that we are having. And the digital banking is perhaps even more than 2 or 3 quarters away, right? So that's how you think about -- we are having it in various stages. They're all coming. And the difference between the past -- the difference between the past and the present on all of this is that the IP in all of this is our IP, right? That is the very fundamental difference. IP is ours. It is co-creation unlike the past where you take the business requirements from various products and channels, give it to our technology who puts the business requirement documents, takes it to a third party, kind of think about a big tech major, right? I think about the [ PCS ] versus kind of big tech major. Go to them to develop and then implement it with the big data tech majors like Oracle, IBM, et cetera, and then you jointly implement it. That's historical approach. But this approach is IP is ours, co-creation with tech start-ups using the latest technology, right? The engineering is very different and on the cloud. So these are the kind of new-age things that are coming through these kind of processes.

Okay. Srini, I've got 2 bullet kind of questions. Is that okay? These are not long questions. Do you have another 2, 3 minutes? We are running out of time. So...

No, go ahead. Go ahead, Sumeet.

Okay. I don't know whether -- so these are very pointed questions. I don't know whether you have the data or not. I mean how many PIN code across India do we currently issue new-to-bank customer -- cards to new-to-bank customers? And how do you think of expanding reach across more PIN codes?

No, this is very proprietary in terms of our credit algorithm are defined. But I can tell you customer from the northernmost, to easternmost, to westernmost to southernmost can apply. It is available. Our credit algorithms will decide which segment, geography segment, customer segment, income segment, salary segment and so on and so forth that they offer. And when they offer, first, whether they will accept or reject. If they accept, at what kind of a credit limit they will accept. So each one, it goes through our proprietary credit models in terms of how we decide and do this. And this is a very dynamic process, right, because that is part of the feedback loop of the credit analysis and the credit that feeds into underwriting and, thereby, into the origination that is a continuous loop that happens, right, in terms of monitoring and making the adjustments and the changes.

Thank you, Srini. And I'll leave you with one last data question. What percent of tech spends are being capitalized? Have you mentioned that anywhere? Can you give some color to that?

What part of?

What percentage of tech spends annually are being capitalized?

We have not published that data, right? We have talked about previously in terms of tech expenses as a -- the tech expenses as a percent of total expenses or tech expenses as a percent of revenue and so on and so forth. But what gets capitalized is not something that we have talked about here.

No worries. So yes, broadly I'm done with all the questions. Thanks a lot, Srini. Thanks a lot for extending your time. Have a good evening. Happy Holi to you and all the viewers. With that, let's end the session. Thanks a lot investors for attending this. If you have any more questions, do reach out to us. Thank you.

Thank you. Thank you all for engaging, and Sumeet, I appreciate leading and getting to the questions that you did. Thank you. Have a great time. Have a great weekend, all of you. Happy Holi to you all.