TMBThanachart Bank Public Company Limited (TTB-R.BK) Earnings Call Transcript & Summary

So good morning to you all, and welcome you to TMB live webcast. So as usual, after we release the financial performance, we will set up the live broadcast for management to discuss with you about the performance of Q3. For this meeting, we have the management with us, Khun Piti, CEO will walk you through the COVID response and the guidance of 2020. And then after that, Khun Prapasiri will walk you through the performance of Q3. Khun Praphan, our President, will walk you through the integration update and after that, we will have the Q&A sessions. So if I will pass this microphone on to Khun Piti.

Good morning guys, and thank -- first of all, thank you for sharing your morning time with us on a Monday. Before we go into the details, which shall be presented by our President and our CFO, I would like to walk you through the page I've called debt relief profile. We'd like -- first of all, I would like to share some light on the debt relief in Thailand because even the people who live in Thailand, analyst who cover Thai banks still very much confused by what happened, what is allowed, what is not allowed. So let me walk you through this a bit. First of all, there is a lot of misunderstanding and misconception about debt relief and the policy cliff that would happen in October. Maybe I would like to walk you through from the bottom of the page first, the retail customer. At the very peak, retail customer will see, for debt relief, reached 45%. And for retail, they have to call to opt in. It's not automatic that everyone can get except the unsecured loan, where bank of Thailand changed the minimum payment from, let's say, 10% to down to 5% or 3% of the monthly installment. So basically, everyone under unsecured, depending on which type unsecured get the [ umbrella ] debt relief and still up until today. They will change it back about 1, 2 years down the road to bring it back to the normal level. That's why the unsecured has much lower debt roughly at the beginning and can come back strongest because not that customer are getting healthier, but they get -- already got their broad best debt relief by reducing minimum payment per month. Next one would be the mortgage. As you may know, in Thailand, repossessed cars is much easier than hold. So customers who have problem with cash flow, and they have to reprioritize that cash flow, they tend to pay high-yield first and pay home car loan first because that's where car can be repossessed or they can avoid paying high interest on unsecured. But home loan would be the last because repossession process in Thailand is quite long. So you cannot [ wake ] people easily. However, we see that the customer who still in debt relief has dropped quite sharply from the peak of 40% down to 15%. The last one in the retail group, the hire purchase, it come back quite strongly. Normal repayment come back to more than 90% level, where 7% still stay for debt -- some kind of small restructure to the maximum of skip payment for another 3 months. That would be the picture from the retail. However, we are not too happy or too pleased or too complacent about the situation that TMB and Thanachart is among the lowest in terms of customer remain in debt relief of around only 10% for the retail. Because the global economy and Thai companies still weak, the 90% on the left could start to fall into 10% on the light blue if more unemployment kick in or those who are still paying under 90%. They already lost their job, but they've paid their monthly installment of the cash reserves. So we have to see what's going to happen in quarter 4 and next year. Back to the upper part, the commercial customer. As you may know, the debt relief in Thailand come in multiple phases, Bank of Thailand urged Thai banks to give debt relief to retail customer first. And when COVID seemed to last longer, then we expect Bank of Thailand would like Thai bank to help SME first through the process called automatic opt-in, meaning that if you do not want to get debt relief, you have to call bank to opt out. That's why all the SME can get that moratorium right away. That's why on the upper part of the graph, you can see that majority up to 99% of the customer in SME actively decide to opt-in because everyone can get it without seeking from the bank to approve them the debt moratorium. But for corporate, corporate have to call banks if they would like the bank to help them to extend or to provide them with a debt relief. The trend is coming down but percentage is going up because in Q3, we decided to trim our loan book, exit the industry that we think that it has not so good outlook when we can still exit them. That's why percentage-wise, it's increasing, but amount-wise, it is coming down. We will start to see [ clearer ] -- clearly picture in Q4 and Q1. Why? Because retail -- all retail customers will come out of debt relief period by Q4, starting from October, that all the portfolio will come to the end of debt relief. However, for the SME, they will start to end that debt relief period by October. And if definition of stage 3 is 3 months, 90 days, then you will start to see NPL by Q1 next year. Bank of Thailand allow banks to look on a case-by-case basis which customer that they still believe that they have the hope that they can come back and then bank can do preemptive restructuring. And with that preemptive restructuring, Bank of Thailand will allow them to treat as stage 1. However, we do not take that lightly, meaning that we do not use fully the forbearance exempt from Bank of Thailand for provision calculation. We adjust our own ECL model. We start to put more and more overlay on to customers that we believe that they become less stronger or start to have problems. So the stronger the medicine that customer would like the bank to help, the more provision we will put in. And the CFO will walk you through the provision that we set aside in Q3, meaning that the trend will continue. So I would like to go to the next page on the guidance and our strategy and our view. We believe that this is not a good time to grow our loan book. We have to be very careful to grow on a very selective basis, meaning that we want to shrink our portfolio a bit to make sure that our portfolio is healthy and well optimized because deposit is growing in -- going in very strongly. Thai customers really worry about situation. So they risk their own investment, they sell mutual fund, they sale risky asset and put into bank deposit. So our deposit coming in sharply. So we use this as opportunity to also optimize our costs by putting out no franchise or low-franchise deposit from commercial bank to make room for retail deposit to grow -- and also, we adjust our flagship product, no fixed rate by providing good rate only to the main bank customer, and that's how we can manage to -- our deposit growth. Otherwise, the deficit will be too huge, and then we will have pressure to grow loan. That's how we can manage to keep NIM in quite a good shape even the multiple rate cuts from the beginning of the year. Also, CFO will explain that the NIM will not be this strong in Q4 because we get some benefit from the way we calculate EIR due to our merger at the end of last year that we have to calculate differently from other peer banks. For the fee income, it's going to be a tough year for Thai banks, particularly for us TMB and Thanachart. On TMB side, our fee income mostly come from insurance and -- bancassurance and mutual fund. On the bancassurance side, because insurance companies have to reduce the return from the saving product, insurance that's very much saving product from the shelf because what they can invest and get return from risky assets become much lower than they have to also reduce the risk appetite of that portfolio as well. So customers are adjusting into the new normal, the much lower rate in the new environment. And for the mutual fund part, Thai customer derisked a lot. In the past, we will get benefit when customer would like to buy a risky asset through foreign investment fund, and we can get front-end fee but now the mode is very much risk-off. So they only buy a low-risk asset where our fee is much lower. What we did well is the cost-to-income. During the past 3 quarters, we can managed to not only not hiring new staff so that we can move people around, redeploy to the area that we still need people. And we use normal attrition and early retirement that we offer in quarter 2. So far, we can reduce almost 2,000 people, and we can manage. By year-end, we should be able to close up to 100 branches through co-location branch. That's how we manage the cost. On stage 3, we can manage to keep it at low level because we write off, we continue to sell that asset and NPL to make room for Q4 and Q1 next year because we believe that this is just the beginning. It's a good beginning for us because retail customer come back quite strongly, stronger than we expect on getting our debt relief, but we cannot be too complacent, too pleased about the situation that we successfully managed because I think that Q4 and Q1 will be a tough time for everyone. So on credit costs, even today, we are on the low side. On Q4, we will put more overlay to prepare for -- yes, we hope for the best, but we must prepare for the worst. If COVID will last longer than we expect and global economy is still weak meaning SME will not come -- we will not be able to come out from debt relief as strong as the retail. And retail itself, we may start to have more and more unemployment issues in Thailand, where Thailand always benefit from low unemployment rate, so let's see. But what we need to do is to prepare for that situation and provide more overlay on to the provision that we already have at quite high level. So may I pass to CFO, to walk you through the detail, and we can come back to Q&A.

Good morning, everyone. Maybe I will start from Page 7. What we discussed since we do the business combination with Thanachart and TMB, we focus on balance sheet optimizations, and the plans worked well during this past 3 quarters. You will see that on the total loan, Q-on-Q, the balance has been reducing by 1.4% negative. This is because we continue to reduce our low yield and high risk using the risk-adjusted pricing. In terms of the commercial loans, the decline in the balance is mainly from the scheduled repayment, so we are aware of that. And we have adjusted our underwriting policies to match with the market sentiment right now. And we continue to focus on the supply chain lending where we have business relationships with the main suppliers and their chains along the line. So by that, we believe that it will help improving the risk profile of the lending that we offer to the commercial sector. In terms of the retail loan portfolio, you see on the bar chart to the right that 52% of the loans are retail-secured loans, mainly home and the car loan. Both home and car loan, even though the balance seems to be flat off, but I am pleased to advise that on -- particularly on the car, we see the 4 consecutive months of increasing in the new loan, particularly for the new car. But the balance has been not improving because, as you have mentioned, that in our retail portfolio, we received repayment back from the customer who in their cost relief. So the payment -- repayment again outweigh the amount of the loan that we can grow and we grow on the risk-adjusted pricing. On this slide, you see that our SME portfolio is only 7% of the entire portfolio. So on to the next slide. On the deposit, as CEO has mentioned, that we do the, what I said opportunistic rundown of the high cost TD and NCD when it matures. And we replace our funding with hybrid deposits. You see that it's 48 -- 49% of the deposit. That is from the hybrid means that No-Fixed, ME and Ultra Saving. We still see strong franchise in this deposit domain. This deposit balance increased 7% Q-on-Q or THB 42 billion. We see new customers coming. And each of the customer, we see higher ticket size per customer. Also, we have been doing very well in the grouping of the CASA, which the flagship is of the deposits. Because we have relaunched new customer value propositions, we observed that Thai people like protections. So we offer a simple and easy insurance coverage necessary to the target group of customers so that they can have the protection at the price that they individually could not be accessed to this pricing. So we -- this new customer, CDP, on the accident protection and life protection has reached out and appreciate by the target customer group, and the momentum is increasing. On the next slide, we have been -- since Q4 last year and Q1 of this year, we talk about the directions on the M&A. But when Q2, we see the crisis. So we shift our focus on managing the liquidity. You will see in May that our deposit has peak. Therefore, what we have to do next is manage our excess liquidity to avoid negative carry because of the customer has been on the risk-off mode. You will see on the cost of deposit line chart that our cost of deposit has been reducing 55 basis points from December last year to this quarter. And net interbank balance has been at the level that we are very happy. What we did, and maybe we can turn on to the next slide, is that we -- our liquid assets excess -- exceed THB 300 billion, and we have a strong and stable funding. You will see on the graph to the top right that the contributor to our funding base is still retail deposit customer, which is on the increasing commercial deposits on the declining and interbank borrowing is on the declining. What we did is that we use the data analytics to understand our customer behavior. So we offer the right CDP which match the customer needs. So our cost of deposits decreased while the deposit balance increased in the segment that we want the relationship with. And on this slide, our LDR is 96.6% and our LCR on the bottom left stands at 199% LCR. So on the next slide, maybe we do from the far right first. Our NIM, as CEO has mentioned, the NIM has been improving 4 basis points Q-on-Q even though the loan yield on the middle chart declined 5 basis points Q-on-Q. But the increase in the NIMs, majority is contributed by the reduction in the deposit cost, which decreased 20 basis points Q-on-Q. Now let's talk about the orange bullet points into the loan yield. Maybe I have to refer back on this one. If you remember in December, when we acquire hire purchase portfolio, we must have -- we have to do the fair values of the hire purchase portfolio of TBANK at the -- because of under the business combination accounting rules. So we observe that there is a fair value. So means that we are acquiring this hire purchase portfolio at free NIM. What we need to do, and we have been discussing since Q4 last year and beginning of this year is that we have to amortize this premium since it will -- over the period of the hire purchase term and the portfolio term at the acquisition accounting point of time is approximately 6.5 years. So we have to amortize that premium over the 6.5 years. Where did we account for this one? It will be the expense and impact to the hire purchase yield. What happened this year is that since Q2, we observed -- well, during our exercise of fair value review, we observed that the customer in hire purchase portfolio, as CEO has mentioned, big chunk of them sought relief. So they have to extend the payment for approximately 4 to 6 months. What impact to the accounting treatment is that we have to review this. So the extent the portfolio lifetime has been extended by the period of the skip payment extension, therefore, the amortization curve will be flatter. So we wait and observe until Q3 to ensure that this is the pattern for skip payment. And therefore, we have to adjust this accounting impact in the end of this quarter. This is a cumulative adjustment for the 3 quarters cumulatively. What -- going forward -- and maybe the difference between this adjustment in this quarter is approximately 13 basis points different. So without this adjustment, the loan yield will be 7 -- 4.7%. And forward looking, this expense amortizations will be at the lower graph. So it will have a slightly impact into the loan yield, but not as big as this one. This is -- I'm sorry, it's complicating accounting so I spent some time with you on this one. So maybe it's time to move on to the next slide. We're talking about our fee income. The non-NII Q-on-Q is declining. This is mainly from the noncore investment particularly on the OCI and investment portfolio that we have some gain in Q2. But our core revenue increased 22% Q-on-Q. That's -- the details are on the graph on the right. The main contributor to the increase in core revenue is the mutual fund fee. In Q2, we also gave relief to our MF customer who invest with us. We gave them front-end fee waiver. But in Q3, everything is resumed as normal. So you see that 67% Q-on-Q fee increase in mutual fund. In terms of the bancassurance, we also see an increase in the BA auto fee because of the new booking that we managed to book onto our accounts. Nevertheless, the retail BA fee have shifted from the saving to the protections policy as CEO has mentioned. On the loan-related fee, we also see the 20% Q-on-Q credit fee increased. Of course, on the trade finance and FX because of the economic situations, we saw a slight decrease of 2%. The LG fee is flat Q-on-Q. Move on to the next slide. We have talked a lot about it. Total operating income in short, Q-on-Q, is minus 2%, but let's focus on the total operating expense that we have, reducing total operating expense by 4% Q-on-Q even with the integration expenses. Maybe I walk you through using the next slide. Due to the directions from CEO that we have to self-fund our spending, so we have to explore saving opportunity. It's good that during the integrations, we have the time to review the target operating models. We use the volume base a negotiation point. Hence, we increased the network, gain on operational costs by some significance. So the cost saving of the 3-year cumulative is THB 1.5 billion in aggregate. Apart from the networking and operational cost savings, we also see the contribution from HR cost saving and marketing cost saving. Over the 3 quarters, we are seeing a declining of our head count by 2,000 head count, 500 of that from the ER self-voluntary package that we offer to them. Hence, we have the integration cost of THB 600 million for year-to-date cost. And that's mainly from the ER package and the staff retention that we have offered to retain our talent who support with the integrations and also partly from the branch lateralization, the colocation expenses that we have to pay. We are seeing that the integration expense may increase in Q4 because we have to do a few more things. We are not moving our target integrations which will be in the middle of next year. So this slide presenting cost-to-income ratio with -- or the integration cost, C/I ratio is 46%. But excluding the PPA, the amortization of expenses, I've mentioned, the cost/income ratio is 43%. Next slide, please. On Q-on-Q, our PPOP is flat but year-on-year, it's 4% increase. Let's move on to the asset quality. In this third quarter, as CEO has mentioned, we continue to de-risk our portfolio. You will see on the stage 3 that the balance of the stage 3 has been reducing because of the sales and write-off exercise, we do $2 billion each roughly for the sales and written off. Hence, the stage 3 stands at 2.33%. You will also observe that the stage 2 -- sorry, stage 2 increase -- percentage increase from 7.8% to 8.3%. This is because of our intentional -- intention qualitative downgrade of the customer. They are not fall, but we -- intention migrate them from stage 1 to stage 2 to prepare headrooms for the post-relief customer. So that's about this slide that I would like to share. On the next slide, on risk cost, Q3 risk cost stands at 122 basis points. This is from the ECL model. But nevertheless, we have include the management overlay of approximately THB 2.7 billion in Q3. Hence, our credit cost is 199 basis points in Q3. Maybe we go to the far right, a graph. Our provisions stand at THB 48 billion, and the coverage ratio stands at 132% increase from 114% last quarter. Let me share with you a little bit about our model. So often I've received questions whether we update our ECL model or not. We have updated ECL model, but the data is still from before, before COVID. But what we saw from the updated model is that significant shift in credit risks, the secure that has worked very well. We did not use the local option on the secure, we use international model. So what we see from the model, is that the model catch the customer, who the model observe that no payments coming in, [ they chief the PD ]. So there is a PD chief within the state. Hence, the ECL model has picked up the forecast risk costs. Whether we have updated the GDP into the model, we did not because we would like to keep the model for the long term. But what we did is that we zoom into our customer group. We group them in the green, yellow and red by industry segment and try to understand the impact of COVID to the business. And we then improve -- increase the PD of each customer by the group that we observe. I -- we believe that doing this details analysis, zooming into our customer profile is better than adjusting the model because we do the segment of [ one ], we view them. There are questions also that this year, the accrued interest in the past, in last year, accrued interest will be stopped and reversed. But this year, the accrued interest under the IFRS 9 continues. So we also use the same parameter that I mentioned to put aside additional provision for the accrued interest. For the Stage 3 customer accrued interest, we put 100% of them. And we look -- use the forward-looking model to put more provision, both in terms of the principal and the accrued interest. Okay. Let's move on. So the net profit Q-on-Q has reduced significantly because of the intention to put more provisions asset management overlay for the Q4 and Q1 of next year. Last slide for me would be on the capital. You will see that our CET1 stands at 13.8%. And last year, because of the integrations, business combinations, we have issued the AT1 last year. And hence, our Tier 1 stands at 14.8%. And And our CAR ratio, because of the sub-debt that we have, stand at 18.9%. This is above the regulatory requirement. So that ends my part of the presentation. So I hand over to President kha.

Good morning, everyone, khrap. Today, I'd like to update the integrations that's after we pass 9-month 3 quarter. First thing is product and service. Now we have the 79 co-location branch as of last month, September. Again, co-location branch is to being -- to [ bring ] branch into the one location to let our customer get a good product from the broad bank. We plan to have the 100 co-locations branch at the end of this year. Next, the cash your car or the car loan refinancing. That is our flagship product, has been offered to our customer on both bank channel. That means the Telesales brand and online channel with the referral program model. The next one is on TBANK bill payment. And we already passed with all TMB channels. That is the Branch, ATM and by banking, we call Touch. The next one our new booking small business loan from both bank's passbook at the Merged Bank and have available service in the mobile banking as well. The last one of the product and service is the deposit side. Because the product feature from the both bank has been aligned already that means [indiscernible] [ provision ] same for Merged Bank or for the customer. The second main thing for the staff migration. From the beginning of this year until this month, October, we already done for 700 TBANK staff to the Merged Bank, that's TMB. That we plan to transfer 3,000 more TBANK staff in January next year. Most is our front office, [indiscernible]. The last one for the facility management, we have consolidated our main office building that we move our staff from the MBK to the TMB headquarter, this building, and move our staff of the contact center from the AIA building to the Wongsawang building also on [indiscernible]. That is our integration update for this quarter khrap.

Thank you, Khun Praphan. Now we would like to open the Q&A session. [Operator Instructions] So while we're waiting, we already received 2 questions. Can management share the view on dividend policy? We believe that Thai bank are now finishing the straight test with BOT? Any update on annual dividend payout?

Not really. I think after BOT hear out the situation and the plan, they will come back with the direction again. But I think all the Thai banks, the capital looks strong. However, as I mentioned, the real situation is remained to be seen, and the P&L remain to be tested on Q4 and Q1. Then we have to wait and see until the end of Q4 because dividend -- decision will be made by sometime early next year anyway.

All right. We received the palm from Khun [ Siward Ludin ] Mr. Siward, could you open your microphone, please?

Yes. I yes. So I have 2 questions. First, on your loans under moratorium. So what is the expected probability of default for your current loans under moratorium? So what's left on your book? And for the loans that just went out of the moratorium period for the retail loans. And can you -- since you mentioned earlier about the probability of default, can you give us the updated loss given default for these loans under moratorium 2? And how does it compare versus your previous historical numbers? And the second question is on the OpEx. So the decline in your OpEx is quite significant, excluding integration costs, do you think this is the bottom? Or do you still expect more cost cuts in the future?

Okay. Let me -- it is good question, but very difficult to answer to predict PD and LGD. Maybe I can share the situation and we have to model scenario. We cannot pinpoint 1 number, and that's how we plan the uncertainty. You can see that retail have come out quite strongly. But as mentioned, we cannot be too complacent about situation. That's why we put overlay on top of various breakdown of customer. The things that remain in the Pandora box, I put that way, to be open is the commercial customer. So what we did is we call every individual customer about a month before to check what would their will and what kind of help they would need from the bank after the end of debt moratorium period. Around 80% of the customer, the SME customers, say that they can come back and pay as normal. About 5% say that they will be in very bad shape and need total restructuring and big help from the bank. The rest of around 15%, they say that they need some small help. For example, the accrued interest during the past 3 or 6 months, they cannot pay back in full in one go. Can we restructure it into term loan, so that they can gradually pay down during next 1 or 2 years, something like that. That have to be tested in Q4. Because if I were those customers, I want to say good thing rather than bad things because customer would worry about bank freezing their limit or cutting their limit if they show that they cannot survive. So that number or that response would be on the positive or optimist side of the spectrum, then it would be difficult to predict PD. LGD-wise is even more difficult to predict because banks through Bank of Thailand will not foreclose the asset too early. On top of that, it ends up having positive impact because when bank slow down or reduced appetite on granting new limit, for example, on car loans, the new car sales coming down, meaning that the supply on the second-hand car will be less entering to the market, that helped put up the second hand price, so we do not -- so far do not see flat of repossess home or car that is sold on to the market and worsen the situation. To the point that the liquidation of the repossessed car, the used car, is a little better than we expect and a little better than the normal situation to deploy that. Now according to Khun Praphan, we have lowest point of repossessed car in the inventory. But again, I don't want to give you guys the rosy picture that things are all right. I do not believe in that because Thailand heavily rely on tourist and export. When that 2 engines really slowed down to the point that it is almost nothing on the tourism side, we will have impact. And what we -- the bank need to do is to build our own model based on the development of the situation, and that is the reason that we put high overlay in Q3, and we will continue to do so in Q4. On the OpEx, what we aim is to self-fund, and we end up positive through the 3 quarters of almost THB 1 billion. The trend will not continue for Q4 because the integration cost would rise, and we will not have big cost saving in Q4. It's cumulative. The cost saving is cumulative. Everything we will do, it come little by little, like closing the branch, reducing the staff. So it's not one-off big saving. But one-off spending will start to come at the last quarter of the integration exercise, which is the Q4 and the next 2 quarters before we achieve the EBT around the middle of next year. But we will continue to manage costs very tightly to deliver the promise that we will use our cost saving to fund the integration costs.

The next question is from [ Sel ]. Fee recovery quarter-over-quarter was pretty strong. Do you expect the recovery rate to be sustainable? If you could [ give ] more detail on fee by segments, by the flagship product? And the second question is about the PPA impact that CFO explained. The lower -- excluding the impact of PPA, was it driven by the lower amortization, as you explained earlier?

Okay, maybe I'll start off with the fee side. Recover, if it means that go back to Q1 level, it's going to be tough, as mentioned. On the mutual fund side, customers are still very much on the risk-off mode. So I do not think that we will get a good front-end fee from selling higher risk product to customers. And that's not -- should not be the way that we put customer to take too much risk during this period. It will destroy our long-term franchise. On the bancassurance, I think it will recover. But again, may not be as strong as Q1 as well because Thai consumer, when they buy bancassurance product, it's still more on the savings side of bancassurance rather than the protection side. The knowledge is developing. I think, Thai people, Thai consumer, getting more and more educated. And thanks to with COVID, people start to plan more of -- to protect their uncertainty, so they start to buy more protection product. However, the big chunk still come from saving products. At lower rate environment, customers become hesitant because they still believe that the rate may go up in the future, and they should not lock in at a low rate. But Q4 will be stronger than Q3 for BA, and overall fee income should be at least as strong as Q3.

So I'll do on the PPA amortization. Let me use a simple -- hope I will get it around this time around. So when -- according to the accounting guideline, we need to do the fair value of the hire purchase portfolio at the date of the business combination. So there is this premium amount, for example, THB 100 million. And that would have to be amortized over the channels of the aggregate hire purchase portfolio at that point in time. So we saw the span of 6.5 years. This is roughly figures. But now those customers because they -- 50% of them, this is rough high-level figures, ask for [ debt relief ] whereby they extended -- they skipped the payment 4 to 6 months, so we use -- we map with that customer and review what would be the revised channels of the portfolio at the business combinations, it's extended to 7-plus years. Therefore, this premium will have to be amortized -- the same amount of premium amortized. Therefore, the amortization cost will be reduced for this entire portfolio. But we have to look back on the 3 consecutive quarter on how much has the impacts has taken in. So we did adjust into the Q3. But forward-looking, that extended period, we'll give a lower amortization as expense net into the hire purchase yield. I hope this helps clarifying things a little bit.

Thank you, kha, Khun Prapasiri. The next question related to the COVID-19 debt relief program. Management mentions about the policy cliff and that we will have to see on the Q4 and the first quarter next year. What would be the manageable NPL formations for TMB Bank to handle after the 4 brands [ can ] or in other words, fourth quarters and the first quarter of next year?

Okay. If we go back in time, the NPL flow for Q1 this year, where every time that we meet, I will keep saying what the risk for the past 2 years. Because I don't -- I'm not so optimistic about the global economy and Thai economy 2 years back. So what we have been doing is to write-off to bring down stage 3 level. So back in Q1, we did that. So our risk cost in Q1 was quite high at almost -- NPL flow was quite high around close to THB 14 billion because we want to bring down stage 2 to make room if customer -- good customer to fall into stage 2. That was the plan, right? But in Q2, we were hit by COVID, and we allow to take preemptive restructuring forbearance by the Central Bank. So Q2 risk costs came down quite sharply to around THB 7 billion. And Q3 also come down to around THB 6 billion. That's why we put overlay on top because we do not think that, that would be the real number. In Q4, the upcoming Q4, BOT still allows for a bank to do preemptive restructuring and keep risk cost or NPL flow at low level. But at TMB and Thanachart, we do not want to put the problems further down the road. So in Q4, what we plan to do is we will let the weak customer fall even we will continue to support them with restructuring. But we all know that many of those do not survive at the end, anyway. So even we help them out by extending the repayment, reducing the interest rate, allow them to take more skip payment, we have to take the loss through NPL flow and provisioning. So we expect that in Q4, the NPL flow would go up close to Q1 level, where we took quite an extreme measure back then to prepare for the worst, the slowdown in world economy and Thai economy. So Q4 NPL flow will not be as good, meaning as low as Q2 and Q3 because we will start to take flow as per the real situation that we expect. Do not fully use forbearance by Central Bank because many SME customers, we will start to show their real result and many retail customers, we will start to show -- most retail customer, we will start to show their real result.

Seem like there's no other questions. So I guess, we will conclude it here. Khun Piti, would you like to say anything before we end the session?

Always appreciate your good questions and your time that you shared with us. I think, you and Thai banks are in a very challenging period. But looking through our peers' performance, including us, I think Thai bank is still in very good shape. We have strong capital, and the many banks taking very conservative approach, even risk, more hybrid capital to prepare for the worst, and we get a lot of support from Central Bank and government. So even challenging situation is ahead of us, but I think we will be doing fine. And thank you for your support and interest always. Having a good day and good week. Thank you.

Thank you.