DBS Group Holdings Ltd (D05) Earnings Call Transcript & Summary

Ladies and gentlemen, welcome to the Q&A session for Third Quarter 2022 DBS Analyst Briefing. I will now hand the session to Michael to begin today's session. Michael, please begin.

Thanks, Diana. Thanks, everyone, for dialing in and welcome to the session for buy and sell side. You've heard briefings from Piyush and Sok Hui. So we can go straight to Q&A. Diana, could you please open the floor for questions?

[Operator Instructions] We have the first question from Aakash Rawat, UBS. Aakash, I believe that you might be on mute. We are not hearing you. Okay. While we wait for Aakash to get connected, I'll put through the next question from Nicholas Teh, Credit Suisse.

Just a few for me. I want to ask on the NIM. You talked about in your scenario that you could reach 2.25% in middle of 2023. I just want to understand, do you expect it then in the second half to start coming off because of funding costs catching up? The other question is on credit cost. You talked about the SP getting to through-the-cycle averages next year. Granted things are very volatile, but I just want to understand how we should think about that big GP overlay that you have and how willing you would be to use that to kind of stabilize overall credit cost? And then the third question is just on dividend. You are looking ROEs at over 15%, loan growth slowing, CET1 above your target range. So implies you could pay a much higher dividend than you have now and still be building capital. So I just want to get your thoughts on how to think about the dividend?

Okay. Because there are 3 questions, the first, on the NIM, it's very hard to definitely figure what the back end of next year might be. What you said is correct. I think deposit rates will continue, the cost of funding might continue to go up, but even that will start moderating. We are seeing a pickup now. Our deposit betas till the end of the third quarter were sub-25%. By year-end, I think they'll get into the 30s, close to 35% next year, by middle of the year, I think they will get up into the mid-40s. I think they'll probably start leveling off at that point. Like I said earlier, we have some tailwinds and headwinds. The tailwinds come from the repricing fixed rate loan book. This $180 billion, some 1/4 of that effectively reprices next year, another 1/4 reprices the year after that. So that obviously gives us some pickup as that repricing happens. But the other part of this, the real drag for us is like for many other people is the treasury book. So the funding cost -- the treasury book is funded in the market and so there is a drag on the treasury book from funding. Right now, this -- we bought it at 4.75%. Now, [ power ] is looking more like 5% on yesterday's statement. And our own sense is, if we get to those levels, you will probably come and plateau at around these levels, 2.20%, 2.25% levels at 4.75%, and we might stick around that level for some time, I think. Your second question on credit cost. Look, it's really hard. All we're trying to do is make sure that we have, what, Jamie Dimon sometimes got a fortress balance sheet just because of the complete uncertainty in the environment in a 5% interest rate scenario. And therefore, like I said, our SP this year is only 8 basis points. If you ask me, based on a bottoms-up view, I don't see them increasing. I'm not seeing stress anywhere, but when I sit back and take a top-down view and say what happens at 5% interest rates, in terms of likely recession, likely slowdown, likely jump in debt servicing, you have to expect that there will be some impact of that when you look down top-down. And so, in my own planning and budgeting for next year, I'd say, let's just go back to assuming that we hit the 20 basis point kind of SP, which we assume as a through-cycle number. Now, on GP, you are correct, we have a lot of cushion. We have $3.9 billion. Like, again, I said on the call, in the first 3 quarters, the model GP number has been reducing. It reduced by $150-odd million in the first half of the year and it reduced by another $200 million in this because the model GP also reflects the underlying portfolio. Underlying portfolio has been continuing to improve and therefore, the model GP has been coming off. So we just took the opportunity to add the overlay to take the GP level back to where we were, at $3.9 billion. Now, what that means is that, our total overlays are about $2.1 billion. We have about $1.8 billion is the model number and SGD 2.1 billion is the overlay number. And to your question, yes, the idea behind the overlay is exactly that, that we will use it to -- and release the overlay if we need to be able to moderate our provisioning level. Frankly, if the world looks much better than we think it is right now, we will probably release it irrespective. If the world looks worse, then this extra cushion that we have allows us to moderate and manage the income more sensitively next year. The last question on dividend. On the dividend, I noticed a couple of people said, why is dividends so low. Frankly, I have been quite clear. I've been saying this from the second quarter itself that we recognize we have a lot of capital, we recognize that we have the capacity to pay more, but it is something that we'll evaluate at the end of the year. And so, it is not something that we are planning to do in the third quarter. So I think perhaps for some reason, it is a misleading of an statement of intent. We will look at it at the end of the -- if you look at our history, we've always done our dividend moves, by and large, at the end of the year and that's what we will look at. Like I said, we've -- all the things you've mentioned are correct. Our capital rates are high. We're above management operating range. We actually -- if we go into Basel IV, we create even more capital and our loan growth outlook is not humongous. So we do have the capacity to return money to shareholders. It's a timely question really.

Our next question is from Jayden Vantarakis, Macquarie.

Can you hear me okay?

Yes.

Great. Well done on a really strong set of numbers. My question is on the same topics as Nick's. So forgive me if I am just going into some more detail on a couple of them. Previously you had a really helpful NIM sensitivity where you said SGD 18 million to SGD 20 million of revenue for each basis point. And I understand there's some non-linearities, but can you update us on what that might be in the current sort of rate environment and what that new sensitivity may be, if you're able to? The second question I had on credit risk is, you must have done some sort of scenario analysis in terms of what the outlook may look like. What would you say a bear case could be for that credit charge? Those are my 2 questions.

So on the first, you're right, it's non-linear and that's part of the problem. So far till the end of September, the SGD 18 million to SGD 20 million has been holding despite the bigger drag on treasury, but as you look out, that number changes. I think one way to think about it, and it's not linear, but let's say one way to think about it, is that, whatever we've been able to lock in so far is locked in and that from here on, rates go up, the sensitivity drops to about half. So from SGD 18 million to SGD 20 million, it drops to about SGD 10 million per basis point, but for the delta, not for the drop. That's really not how it works. That's an easy way to model, and I figured it's a way for me to understand, but in reality, what happens is, even your stock reprices, but the delta doesn't actually -- it's not like in the delta, you pay up so much, but that's a good way to model, if you want to figure out the model. Different way to model is this, deposit beta [indiscernible]. Our deposit beta was 23-odd percentage so far. I think by year-end, with what's happening in the environment, that deposit beta is probably going to get closer to 35% -- from 25% to 35%, and then our assumption right now is if you look into next year, by summer next year, it might get up to 45%. If you triangulate the [ 9, 10 ] for the delta, and this deposit beta, [ it just ] give you about the same number. So that's the 2 different ways of thinking about where this might wind up. When you put either of those prisms and you work it back into NIM, you get that kind of NIM that we're talking about plus-minus 2.25%. So hopefully that's helpful. Your second question was on cost of credit. We are -- [ often the ] scenario planning we're doing right now is not giving us any line of sight to a material deterioration in credit. And that's one of my big challenges. Nobody has seen a 5% interest rate environment. If you look at our book and look at the underlyings of the book, the large corporate book, which is a big chunk of our loan exposures, continues to be very solid. It's very pristine and my own sense is that, even at 5%, we don't see pain in that book. We've talked to a lot of clients, we've talked to sectors, industries. Some of them are benefiting after COVID, but by and large, I think the large corporate book is quite resilient. Then if you look at SME book, that's about a $40 billion book, 10% of our total book. Now, that's the one I'm more uncertain about because at 5% interest rate cost, SME hurt a lot. Having said that, we've stress-tested that book at 6% and 7% and we didn't see a lot of damage. And I think part of that is because, in the last 3, 4 years, we've really refreshed that book a lot with the China issues in the past, the TMT supply chain, the COVID. So that's a fairly seasoned book. So when we do the bottoms-up on that book and build up, even that's not extraordinarily large, $200 million, $300 million, $400 million potential cost of credit when you do it bottoms-up. On the consumer book, the mortgage book itself is also quite solid. Our mortgage, the bulk of it is in Singapore, and as you know, we've always done -- even going into our loans, we've done 3.5% interest rate assumptions. We now do 4% interest rate assumption, and our loan-to-value in the portfolio is very low. It's been -- overall at a portfolio level, it's about 50% loan-to-value. So again, when you do bottoms-up, I don't expect a lot of damage in that book. Again, historically, over the last many decades, we've not seen too much damage in that book. And then there is unsecured consumer book, that's about $10 billion. Half of that is in Singapore, which is actually fairly well controlled. In that book, consumer -- because some of the rates are capped. Credit card rates are capped. In some markets, unsecured loan rates are capped. The consumer doesn't really wind up suffering a lot on direct payment. The bank takes a squeeze on margins, but you have to keep in account that disposable income of consumers will be affected overall in a 5% environment. So you should expect a delinquency pickup in that book. The reason I am going to such detail is when I do this thing bottoms-up, I find it hard to even get to my normalized credit scenario, right, my 17, 20 basis points, I'm not even seeing that. But I'm just being cautious because there are a lot of unknown unknowns in a 5% world. And so, what happens is so uncertain I'd rather be cautious and be prepared for things that we've not foreseen or forecast.

Our next question is from Nick Lord, Morgan Stanley.

I have a couple of questions actually. I'll come back on the credit quality, but just first of all, on that NIM point from a trading book being funded. I think some of the other banks who reported have said that, that then plays through in higher trading income, so the net income benefit is 0. Would that be true for you as well? So as that trading book funding cost goes up, we'd also expect to see your treasury and trading income go up by a similar amount?

So Nick, we saw some of that. If you look at our overall treasury book, our guidance is that, trading income should be about $275 million a quarter. And if we look at this quarter, it's about $265 million or something like that. And there, we saw a drag on the funding cost, but we made it up on the other income line and that's correct. But as you look forward, it's not clear to me it's a one-to-one correlation. I think some of the -- what you give up on the funding cost, you make up on the other income side, but there's obviously volatility in that number. I still think that we should be able to stick to the $275 million a quarter overall guidance. I'm not changing that guidance, but you could see choppiness in that from quarter-to-quarter.

Most probably -- so Nick, maybe since we don't actually show the full performance summary, if you have seen the treasury customer segment income, so the trading component quarter-to-quarter will be quite flat, close to the $265 million for the last quarter. So what you see under the strong other income line, so in addition to this shift between NII and non-NII, which is about $100 million, the other component is actually stronger customer [ bit ] of treasury income. That's what you're actually seeing, the customer treasury income has also been stronger.

Okay, perfect. And then just coming back on credit quality and really following up on Jayden's question, and I think you gave us a pretty clear idea of some of the transmission mechanisms. I mean, to me, it seems like there's just 2 potential risks for you. One would be that interest cover begins to fall because the EBITDA comes under pressure and rates have gone up on the corporate side. And then, I guess, there is a risk around property exposures because presumably as rates go up, property prices start to -- or asset prices starts to come down. So have you thought about those? Can you give us any metrics, for example, on what the interest cover is on your book or what percentage of your book has stressed interest cover or anything like that? And also, if we did start to see property prices moving down materially, especially in the commercial space, are there levels of price decline that would begin to cause problems for you in your book? I'm thinking commercial property rather than residential property here.

Yes. So Nick, I don't have the data on the interest cover, though, obviously, we stress-tested it and we can get somebody to pull that out. But when I said that the large corporate book and this thing is quite pristine because we are not seeing a large number of companies go into this stressed interest cover. Our stress tests are quite [indiscernible] we stressed at $200 oil price, we stressed at high interest rate, we stressed at weak currencies, and only a handful of names frankly that we think run into our, let's say, watch list which captures all of this, is actually -- our total names in our watch list are about 4% to 5% of our portfolio right now. And the watch list names are the ones where we think that interest cover, debt service cover, et cetera, it would be something that we'd have to watch for. So it's not a big number, but I can get somebody to get that specific thing for you. On the property price decline, the reality is our cushions are very good. So apart from the mortgage book, even on the commercial property side, our loan-to-values are generally quite low. And therefore, our stress-testing, both in Hong Kong and Singapore, where a large part of our collateral is commercial property, up to a 30% drop in commercial property prices doesn't really create a problem for us. We still are in the money.

Okay. That's perfect. And then finally just on capital. In terms of mechanisms for returning capital, can you just -- I mean, obviously, your share price is quite high. So I don't know if share buyback sort of work on that basis and there's problems of dividend and dividend sustainability. And, I guess, if you're talking fortress balance sheet and if other people in your sector haven't got fortress balance sheets, and after a year of downturn, you might be in a good position to do M&A. So can you just talk about mechanisms of returning capital and how you might be able to use a strong balance sheet over a downturn scenario to add value longer-term to the group?

So Nick, on the second part of the question first, M&A, we don't see deals in the last couple of years and we're still absorbing them. So LVB like I said, it's integrating well, but we still have to extract value from that deal. So we've got our hands pretty full in India for the next year or 2. Taiwan will only integrate by August next year. So I think we have our hands full with that for a period of time. And we are still consistent with our long-term philosophy. We're not looking at doing an earth-shattering M&A of large-scale because we think that is still quite distracting to do. So I don't anticipate too much on that front. We're always open to bolt-ons. So if in the next year or 2, we get some opportunities, but those -- the largest deal we've done is a $1-odd billion. So those are not going to be huge drags on capital as we look for those opportunities, whether they are a line of business opportunities or fintech-related opportunities, but they are not going to be massive drags on capital. So what do we do with the surplus capital? As you correctly said, we've got effectively 6 choices if we want to return some capital. We can increase our regular dividend. We can pay a special dividend or we can do share buyback. In fact, everything is on the table. As you correctly said, share buyback at our current stock price and valuation might not be the most attractive part, but again I hasten to add, it's not something the Board has reviewed. It is something we will do between now and the end of the year.

Our next question is from hey Yafei Tian, Citigroup.

I have a couple of follow-up questions on net interest margin. The first one is, is it possible to give us an indication of where October or September exit margin is for you? And also I didn't catch on the media call 100% the part of the portfolio that can reprice. Can you just either repeat what you have said in terms of what is the quantum of the book that can reprice over the next 1 or 2 years? So that is on the net interest income side, and I have a question on costs.

So I said that we are likely to end up with a NIM of around 2% sometime between the third and fourth quarter. I think we are currently at very close to 2% at the moment. And that includes the drag from the Treasury Markets NIM.

And on the $180 billion, do you want to give them or give them half [indiscernible]? Do you want to give the specific details of $180 billion, what we paid?

So for the $180 billion, I think they are very sort of -- you're talking about the...

Book that reprices...

The book that reprices. So up to sort of end of next year will be about $60 billion and then in the 2024, I think we'll see another $40 billion over and then the rest reprices after that.

And the nature of the book, like I said, about half of it is loan and I think $80-odd billion, the $80 billion, $90 billion and about $70 billion, $80 billion is securities. So I got -- there is a bunch of security and there is about $30-odd billion of interest rate swaps, is the hedging that we do to build duration. So that's...

So you are talking about the composition of the $180 billion. So $80 billion is in fixed rate loans, $22 billion in fixed rate mortgage, sort of fixed rate -- fixed deposit rate for mortgage, $22 billion. We put our interest rate swaps from about $33 billion to convert floating rate loans to fixed rate and we have a corporate treasury portfolio of $46 billion. That is the composition of the $180 billion.

Understand. Very clear. I have a question on OpEx as well. Clearly, inflation is running high in Singapore. We see that translating into the staff expense increase. So assuming next year, the asset quality environment remains relatively benign and then you benefit from higher margin, just wondering what kind of cost inflation are you thinking about for next year?

So I think you should be expecting what we had in the third quarter, the 9%, 10% kind of cost growth for next year. Just because when inflation is a CPI that is the wage growth will continue to be somewhat stronger.

I don't have a lot of visibility in terms of OpEx trends beyond 2023. So assuming that margin expansion kind of slow down after 2023, can cost growth moderate after that?

Well, the costs are a function of, a, headcount and wage levels and we keep driving efficiency in the business. As I've said before, we have a strategic cost management agenda. We've added a lot of people into technology. Some of them are one-timers. So we've added a lot of people for Lakshmi Vilas and for integrating of Taiwan, 700, 800 tech people, for example. Those will run off and we expect them to rationalize by the end of next year, for example. So yes, we do have some levers to be able to bring costs down. Again, if you look at our history, we've said every time we have the opportunity and we get improvement in margins or tailwinds, we are willing to invest a little bit more in some of these activities, and then when things are slightly tighter, we have continuously been able to discipline our costs and bring them back. So I think we know how to do it.

Our next question is Weldon Sng, HSBC.

Can you hear me?

Yes.

Okay. Can I just ask the NIM sensitivity? I think previously you've guided -- that guidance is roughly about 6 basis points per 25 basis points, that rate hike. Now if we think your -- I think 2.26% guidance at the middle of next year, that implies a 3.5 percentage on the upper bound, which is what -- you've said previously that that is what the sensitivity should last up to. So I'm just trying to understand your 2.25% guidance. Is it that you are just taking the sensitivity up to that point and then saying that from then on it'll be less? Or what -- can you give some color on that one?

So I just gave 2 prisms on how I think of sensitivity, and I don't do the math 6 basis points per 25 basis points. I don't use that prism. So I can't relate to how you think about it. But I think the 2 prisms I gave you get to this thing. One, we look at what is the deposit beta, which means of whatever Fed rate increases that you see, how much you have to pay out for funding costs and deposits. And that we see rising from about 25% to about 35% by year-end and to about 45% by middle of next year. And the other prism I gave you is that, we assume whatever we have right now as locked in at $18 million to $20 million per basis point. Then from here on, whatever rate increases you see, you should effectively take a $9 million, $10 million per basis point on top of that. So both prisms will get you to around the 2.25% number. So you can actually work it back and see what it means from your lens and the way you look at it.

Okay. Sure. And then just on that, is it -- when you say peak at 4.75%, are you seeing that it stays there for the rest of the year in your assumption?

Well, in our assumption, we assume they hit 4.75% and hold it. I don't see the Fed cutting rates next year. So our assumptions are they get there and they hold all of next year. Now, after yesterday, I would bet that power will get to 5% and -- but my own this thing is that, I don't see Fed cutting rates next year, even if you start seeing a slowdown in inflation, my base case is you start seeing cuts only in 2024.

Right, okay. So if you are saying that the rates -- the Fed rates hold, then is it correct that in the second half, there should still be some repricing that -- so the 2.25% is only a midpoint and there is upside to that in the full-year number?

So what we don't know, and somebody asked the question before, I think there will be some tailwind on repricing in the back end, but at the same time, there might be some more marginal deposit repricing also when rates are at 5%. So my current this thing is, it will probably wind up at a plateaued level, at about that level.

Okay.

It also means if Fed hikes rates further, so it takes that 5% or even higher compared to the 4.75%, then the opportunity for NIM to be higher would be sort of lifted as well, right?

Okay. And then I think just -- answered the question, of course. Just -- can I just ask a bit more detail? So just from you're talking about that -- from this third quarter level, which seems to be just a higher base from this last quarter. Is this the new base, in which we should add the 10% addition into next year on a quarterly basis?

No, I think we should look at the 10% growth on a full-year cost basis.

Okay. Right. I see. And I think you talked about inflationary pressures just now, but is that also a lot of bringing forward the investments?

There is some of that. So it is partly wage inflation. Partly it's -- these one-timers on the integration we need to do and partly we have brought forward investments because we have the top line to be able to afford it. Particularly in the tech space, we've tried to do some of that as well.

Our next question is from Harsh Modi, JPMorgan.

A couple of questions. More than a couple actually. First is on, could you explain a bit more on what changes have you made bottom-up in terms of underwriting over, let's say, last 2, 3 years, which gives you comfort that even if things go to -- if we are in a reasonably tough environment, we'll only hit 20, 25 basis point of provisions and not, let's say, a 50 basis point?

So Harsh, you asked me this question the last time as well. You asked me this question twice before, and I've given you a detailed and elaborate responses twice before on why we think our credit underwriting is better. I can redo it for you, if you like, but there are 2, 3 big changes we've done which are giving us value. The first is, we tightened our entire target market on -- including our risk acceptance criteria across the board, both at industry and segment levels. So we reduced our obligor concentration cap, for example. If you look at the 2015-'16, we took some large hits in oil and gas, where we trimmed down all our exposure to those kinds of mid-cap names and we changed our TM racks around that. So there is a material shift. The second thing is, we've dialed up our industry focus. So we had some industry expertise, but we've really built a much stronger industry expertise and we've aligned it all the way down. Earlier our industry expertise was focused on the top-tier companies. Now, both the mid-caps and even the SMEs, all aligned by industry. And so, in the industry we will pay a far more active role in both credit underwriting and portfolio monitoring. So the process itself of how we manage that has changed quite substantially. The second thing that we benefited a lot from is much better use of data and analytics. So we've dialed up massively, not just for underwriting, the algorithms for underwriting, but a large part of our focus has been on early warning detection and portfolio management. And therefore, today actually we -- anecdotally, we today go and tell companies 3 months before they know that they're going to have a problem because our data analytics is actually -- and AI is helping us a lot to stay ahead of the curve in terms of understanding where the weakness might be and what we need to do with that. We got into -- we've also put a lot of technology into the end-to-end process. So we don't have -- we have much better measurement of the issues and much better end of -- I mean, end-to-end capacity to see through what we have and where we are. The other thing we've done in the last [ 4 years ], we started churning our portfolio a lot more activities. So we actually sell down a large part of our exposure instead of keeping them all on our books. So if you look at the collective of that, you can see that showing up in our credit process. The reason why we changed our guidance from what used to be 20 to 25 basis points to 17 to 20 basis points in the last 2, 3, 4 years is that, we're getting increasingly confident that the sum total of all the changes we've made are coming through in the way the business operates.

I know we had that conversation 2 years ago, but a follow-up. So this watch list of 4% to 5% of the book that you alluded to, have you had any discussions with some of these guys saying that you may be at risk? And are you at a point where you have started talking about restructuring some of those higher-risk borrowers?

The higher risks are very small, right, and yes, of course, we talk to them. The biggest part of the 3%, 4%, maybe 3.5% to the 4% is my first category of just monitoring them because of the environmental conditions, right? There is nothing in that numbers or the data which is -- but it's the macro environment thing, but we put them into the watches anyway to take a look at. So we have a [ amber at peak ] this thing. So the ones -- the bulk of it, 90% is probably [ amber ]. So it's only 10% in the next 2 categories and those, of course, we talk to the companies and see what we can do about helping them restructure or stay out of trouble. Yes, we do.

Right. So is it fair to say that we should see your Stage 2 loans moving up over the next quarter or 2?

No, we are not seeing that kind of migration. In fact, if you look at our sort of general provision portfolio improvement, it's been coming down. We said for 9 months, it came down $300 million and that's why we're able to increase our overlay by $300 million. So the underlying portfolio actually has been improving and you're actually seeing upgrades. That's the reason for the portfolio improvement. So it's hard to say next year, but so far, our experience has been quite good.

Right. Okay. Second question is on Hong Kong property exposure. I know you said that even the commercial property exposure in Hong Kong looks so clear and I remember about a decade ago you had stopped doing mortgage in any large size in Hong Kong. But given whatever is your current exposure to Hong Kong property, how do you think about it? How much of that worries you? And it's not the median exposure. It's really more the 90th percentile exposure. And what kind of numbers do you think we are looking at there, if at all in terms of risk?

Our mortgage book is quite small. I think it's $6 billion or $7 billion and the LTV on that book is 28% or something like that, it's up 30% LTV on that book. And it's not been growing in the last year. It's a seasoned book. On the commercial book, I'm trying to remember what our actual exposure number. I don't have it with me. I can come back to you on the actual exposure number, but the bulk of that is the high-end names, the [ Chong Pangs ] and the Sun Hung Kais, et cetera. That's the bulk of our property book. It's generally the borrowers are top-class borrowers. We've done -- actually just last one, which is why I should know the number, I forget it. We've done a complete review of that portfolio and we're really not seeing any problem. The high-end names are very solid. We don't see any problem even with the drop in property prices in that and we have some exposure in the next year, tier 2 developers, if you will, but those are well secured. LTVs are low. So in our stress testing that we just finished last month, we didn't anticipate any problems with Hong Kong property.

Okay. And the last question, Piyush, is on capital generation and distribution. Given that we have a bit of an RWA increase, some mark-to-market and potentially credit migration down the line. Despite a very impressive 15-odd percent ROE, what kind of capital can we generate? And again, distribution, should we think how sacrosanct is 13.5% upper end or not? Or should we think more in terms of payout?

So I don't know if you think in terms of payout and our operating range we always guide is between 12.5% and 13.5%. So you now said 13.5% is sacrosanct. But there is one other thing, which you're not factoring into your calculation, which is Basel IV. So the Basel IV impact on us is very, very material. Come Jan 1, '24, our capital adequacy jumps to 16% because of the Basel IV impact. And therefore, when you think about the capital cushion and buffers we have, it is for between now and end of the decade, we are looking at capital cushion buffers of 2%, 3%, not decimals of percent.

So -- but this MAS, you think, will allow you to payout between now and, let's say, 1 Jan '24 200 basis points of capital?

You are assuming I want to payout 200 basis points of capital, why would I?

Exactly. So if it is at year-end 13.8%, even after all the decline Q-on-Q, you are 30 basis points above, assuming you generate some in fourth quarter and there is going to be a reasonable amount of relief as you said in less than 15 months. So then how do we think about the quantum of capital that you can pay and can you...

But I think that's an inappropriate question and you should think whichever way you want to think. Even my Board hasn't thought about it. So I've told you 3 times. So you can put whatever you want in your assumption. It's not -- I haven't thought about it, my Board hasn't thought about it. You can think whichever way you want to think.

No, no. That's fair, Piyush. But I'm just trying to understand the guidance of 12.5% to 13.5%. Is it -- are you going to stay with that guidance? Is the Board willing to stay with that guidance? Or do you think, given that we're going into a bit of a downturn, that it is fair to assume that you will have a buffer over and above 13.5%?

Let me just say what I just said. I mean, for some reason, you are very hung up because you assume third quarter, you'd get some dividend increases. I don't know where you came up with that idea from. And I have nothing more to add to what I've told you, right? I would tell you by the end of the year, after we are deliberative at the Board, what it is that we are prepared to do and what we think is appropriate.

Our next question is from Aakash, UBS.

Can you hear me this time?

Yes.

Great. Sorry, I had some technical difficulties earlier. The first question I just wanted to follow up on the NIM sensitivity, Piyush. You mentioned about the deposit beta and you obviously suggested savings rates will go up, which is why sensitivity will be lower. But I think the other variable is CASA outflow, right? If CASA outflow accelerates from here, then that sensitivity can further go down. So I'm just trying to understand what your assumptions are on the CASA outflow. But do you think it will continue at the same pace, do you think it becomes slower or faster from here?

Actually, when I give you sensitivity, I factored that in because I can always replace the CASA with FDs. Right? So when I talk about sensitivity, it's not just repricing the CASA. It is CASA moving. If you look at this quarter, we move $30 billion of CASA, we got $30 billion in revenues more or less, right? So we can always -- if I need the funding, I can always price and get the FDs, but it comes at a higher price and therefore, the interest rate sensitivity and the depositing assumes that X amount moves into FDs and we pay a higher rate of FDs.

Okay, got it. So the 25% to 35% to 45% assumes a certain amount of CASA outflow, which was taken?

Moving into FDs, correct.

Second question I have is on, if I look at your ROE guidance, 15%-plus comfortably. I think it still looks very conservative. If I even go with your base assumption, I think the ROE could be a lot higher. So I believe you're baking in some sort of best-case, worst-case scenarios for certain things like provisions, wealth management income. Could you help us understand some of the ranges in your mind around that? Like in a worst-case scenario, what could be the credit cost like? What could your wealth management be like? In this case, you get to 15%. Otherwise, you could be comfortably above 16% as well, I think.

No, what we baked into our thing, I think we're comfortably above 15% is, credit cost went back to the 20 basis points out of SP and some GP on top of that. I think it's a fairly conservative assumption. We might not need that kind of cost of credit, if you will. So I think that's quite clearly an upside opportunity on the numbers that we're looking at and talking about. In terms of growth, we continue to bake in a mid-single-digit growth environment and grow double-digit growth in fee income. Like I said, our wealth income dropped by 20% this year. So if the environment turns around to be positive, there's no reason we grow -- won't get back to previous wealth income levels, right? And frankly, we've got another $20 billion of net new money. So it adds up on top of that. So it could be much stronger, but again, it's going to have to depend on a far more helpful external environment. Right now, we're assuming that the environment doesn't fall off a cliff from here and we see some more this thing, so you can get to a low double-digit growth rate. That's another example of, and if you want a range, it could get worse, I doubt it, we've been conservative, but it couldn't get much better than what we forecasted.

Got it. And on the mechanism of capital payout, you said all options are on the table, including a special dividend. So if I look back, I think DBS has rarely done a special dividend. So is this special dividend more linked to the Basel IV expectations, which only happens in 2024? Or even -- let's say, if Basel IV implementation were to get delayed, would you still consider a special dividend?

Yes. I think we have enough capital without Basel IV as well. By the way, we did do a special dividend 5 years ago. So we -- I think we did $0.50 or something. So it's not like we've never done one. It's not something we normally do is correct. But our current outlook is that, we do have a fair amount of capital even prior to Basel IV.

So then how would you go about thinking between core dividend increase or a special dividend? I'm just trying to understand like which way would you lean and why?

Our core dividend [indiscernible]. I don't want to wind up at a core dividend level that I need to backtrack on. So other than when the regulators asked us not to, our [indiscernible] is to try and keep it consistent. And therefore, we think about core dividend, I don't want to jack it to a level where in a subsequent year or subsequent quarters then I got to go back and rethink that number. So I'd like to keep some cushion on the core dividend side.

Okay. Got it. And just finally, I might have missed these comments earlier. The trading income gains that we had this quarter, could you give some color on like what was driving it?

So I think I will recap what I said earlier. So trading income is quite strong. You see in the numbers. I think we had a 30% over increase. Maybe that's the number that you're looking at. So a few components. One is actually the T&M where the net interest income, where we said there is higher funding costs, is actually reflected in the higher gains on non-interest income. So that's about $100 million switch between NII and non-NII with T&M overall income being flat. Then we also have customer T&M income both from institutional banking sales and consumer banking that forms the larger part of the remaining increase.

But from a debt standpoint, equity has been challenged. Rates and FX has continued to do well.

Our final question is from Tejkiran, White Oak Capital.

This question is about the acquisition of Citi assets in Taiwan. How easy or difficult is it going to be to get the book directly? In the sense, do you have to reach out to all of these customers and get their consent before onboarding them onto the DBS platform? And secondly, once you onboard these customers, it's quite possible that the behavioral scorecard which Citi used and what you will use will be quite different. So how are you going to treat the churn of the customers based on your risk appetite? And, I guess, I'm just looking for if you have any idea of what percentage of the Citi's book do you expect to hold onto at the end of maybe '23 or '24?

Do you remember the resumption...

I don't have the number off-hand, but I think we have got the contract, which allows us to sort of adjust AUM for Citi. It's very motivated to make sure that at the point of transaction and now it's keeping to those numbers.

The concern is after we get the -- I don't have the numbers off-hand, but when we modeled this thing, we actually did 2 things. One, we compared -- we have a lot of Citi people in our this thing, for example, including people from the credit team. So we have a fairly good sense of their -- had a fairly good sense of their underwriting, both the [ A-Score, B-Score ] model. They are very similar and consistent with our risk appetite. Nevertheless, when we do our modeling, we assume -- in fact, I remember, 15% or 20% customer attrition and staff attrition. That's what we typically put into the M&A deals. Again, I think somebody can get back to you on what the exact resumption on those numbers is.

Sure. And regarding consent, is there a legal or regulatory requirement to AXA to get the consent of all these customers before onboarding their...

Not that I remember. No, no. Not that I remember.

Okay. These customers will be getting a DBS credit card in place of the Citi credit card once they've been onboarded, right?

Correct.

Got it. And one final question, if I may, about the 1.5x multiplier that MAS applied on the operational risk capital. Please tell us if -- is that multiplier, has it expired or do you have any visibility regarding that?

It hasn't expired. I think MAS will do a review in the early part of next year. We'll see after that. So my own expectation is that, in the first half of the year it should expire. But again, I can't speak for MAS. MAS is a large -- because they have...

Congratulations on the results.

Ladies and gentlemen, we have now come to the end of the Q&A session. This concludes today's conference call. Thank you for your participation. You may now disconnect.

Thank you.