© 2024 Alpha Spread Limited. All Rights Reserved.
www.alphaspread.com is not operated by a broker, a dealer, or a registered investment adviser. Under no circumstances does any information posted on www.alphaspread.com represent a recommendation to buy or sell a security. The information on this site, and in its related newsletters, is not intended to be, nor does it constitute, investment advice or recommendations. The information on this site is in no way guaranteed for completeness, accuracy or in any other way. In no event shall Alpha Spread Limited be liable to any member, guest or third party for any damages of any kind arising out of the use of any content or other material published or available on www.alphaspread.com, or relating to the use of, or inability to use, www.alphaspread.com or any content, including, without limitation, any investment losses, lost profits, lost opportunity, special, incidental, indirect, consequential or punitive damages. Past performance is a poor indicator of future performance.
You don't have any saved screeners yet
Intrinsic Value is all-important and is the only logical way to evaluate the relative attractiveness of investments and businesses.
Good day, ladies and gentlemen, and welcome to the TCS Group Holding Q3 2019 IFRS Financial Results Investor Call. Today's conference is being recorded. On today's call, we have Oliver Hughes, CEO; Ilya Pisemsky, CFO; and Sergei Pirogov, Head of Corporate Finance. At this time, I would like to turn the conference over to Oliver Hughes. Please go ahead.
Thank you. Good afternoon to everybody and thanks for joining the call. As we approach the end of 2019. I'm pleased to say that we like what we see. Our plan is coming together on practically all fronts in quarter 3 numbers, and this will continue in quarter 4. I'll start with some mobile app stats because driving engagements and growing customer lifetime value through cross-sell is one of the main thrusts of our work in 2019. This will be the lynchpin of our retail system approach and strategy going forward. We're consistently in the top 3 Russian financial apps in App Store and Google Play, giving us over 17 million downloads of our core mobile banking app. When you compare leading global fintechs, we come out well. We have great UX, a voice assistance, rich product, AI-based customized content and lifestyle services, giving us a 4.9 rating, the highest among all fintech plays, which we see. As a result, our daily active users or DAU has grown by 88% year-on-year to 1.5 million. Similarly, MAU or monthly active users has grown by 84% year-on-year to 4.6 million. Much of this growth comes from the additional content, the offers and the lifestyle digital services, of which we've been adding more and more over the last few years. This confirms the hypothesis that we set out to prove: financial players can have a go-to app with high DAU and MAU that drive engagement. This positions us well for our next 3 years of strong growth in a rapidly changing financial services market. I'll now give you an update on how other parts of the business have been performing, starting with our credit business lines. Our credit businesses have been performing very well. Three quarters into 2019, we've already reached our full year target in terms of net loan book growth. This is due to our strong product, multiple acquisition channels and our brand and servicing advantage. Much of the growth came from personal loans, car loans and home equity loans. By the end of the third quarter, the share of non-credit card loans in the loan book reached 38%. Furthermore, secured loans comprised 13% of the total book, illustrating our confident move into secured lending following successful pilots earlier this year. Despite an uptick in cost of risk, the economics of our lending businesses are very strong, and we continue to book highly NPV positive customers in large numbers across all of the credit business lines. Question of asset quality has recently been a recurring theme, as investors try to gauge where we are in the cycle. As I've stated many times over the last few months, we believe that we are now past midpoint and are in the second phase of the credit cycle. This means that going forward, we'll see an increase in cost of risk across the market. But the speed at which our cost of risk will increase is a function of several factors. Firstly, what Tinkoff does. As we move into the second phase, we've taken steps to stay well on top of credit quality, and we feel confident about where we are. Our vintages are stable. Secondly, what the regulator does. As you can see, the CBR is very interventionist and is not afraid to introduce yet more market cooling measures. We believe that their approach is working. And thirdly, what the market does. While we see that the markets are still doing a fair amount of lending volume, we see that growth is decelerating. It's our view that this is more due to the slight deterioration in risk across the market than to PTI regulation. However, we expect the cumulative effect of RWA and PTI to have a cooling effect on growth as well. We see this as a positive. Our cost of risk increased somewhat in quarter 3 to 9.1% from 8.9% in quarter 2. As previously explained, this is mainly due to the softening of one particular metric, roll rates. Please also bear in mind what Ilya explained on our last call. Much of the increase in cost of risk is down to the front-loading effect of IFRS provisioning as we grew the lending books quickly from a low base. You'll see this number come down in quarter 4 due to the reason -- sorry, usual seasonal reasons. And secondly, the measures that we took to tighten up our underwriting midyear. We explained during the SPO and subsequently, that as we approach the end of the year, we will slow down growth of the loan book in order to: one, manage our P&L result; and two, keep well on top of asset quality. This is exactly what we're doing. So please don't be surprised when you see lower growth for quarter 4. I'll now turn to our non-credit business lines. Our non-credit business lines are on track for strong growth and most of the big ones have broken even, as you know. Tinkoff Insurance had another good quarter, showing 33% quarter-on-quarter growth in premiums earned and 2.2x year-on-year growth, thanks to our mix of online and offline distribution channels. Tinkoff Black attracted another 0.5 million customers in the third quarter, reaching a total of 6.3 million customers at the end of September. Just to remind you, Tinkoff Black is one of our main growth drivers, and we've started to unlock the value in our ecosystem as we open more and more cross-sell arterial routes, much of them centered on mass affluent Tinkoff Black customers. Tinkoff Business adapted to the new regulatory reality and has now caught the wind again. It demonstrated steady growth and reached just over 0.5 million accounts at the end of September. Tinkoff Investments is going great guns and added almost 160,000 new accounts in the third quarter with a total of 750,000 brokerage accounts opened by the end of 9 months. Tinkoff Investments' team added several new services to the app, including an AI-based research engine, which helps investors keep track of the financial ratios of their portfolios, an open API for algorithmic traders and an upgraded robo adviser, which analyzes and optimizes portfolios. In October, our new digital asset manager, Tinkoff Capital received its license to manage mutual funds. We also launched the Pulse social network, which allows customers to meet traders, follow other investors and share trading ideas on what's moving. And finally, we're continuing to devote a lot of time to develop other lifestyle services. Just to remind you quickly, in one click, Tinkoff customers can buy the full range of travel services, including tickets, hotels, car rentals and packaged stores, as well as entertainment services, such as restaurant bookings, cinema, theater, sport, concert tickets and also enjoy various shopping experiences. And we drive content through Tinkoff Stories and Tinkoff Journal. As a result of our growth and focus on the bottom line this year, the group made a net income of RUB 25.1 billion for the first 9 months of 2019. This represents a year-on-year increase of 34% and an ROE of 59% for the 3 quarters. Other noteworthy events were: Markswebb named Tinkoff the indisputable leader of Internet banking in 2019, and Global Finance named Tinkoff the Best Consumer Digital Bank in Central and Eastern Europe. The launch of CloudTips a QR based online service for tipping and charity donations. The launch of Tinkoff [ AZS with BP ], a mono app for payments at the pump without leaving the car. In October, Fitch Ratings agency upgraded Tinkoff Bank's ratings from BB- to BB with a stable outlook. Tinkoff signed sponsorship contract with Russia's top-ranked tennis player, Daniil Medvedev to be our brand ambassador. I'm also pleased to announce that on 28th of October, Tinkoff GDRs began trading on the Moscow Exchange. With that, I'll hand over to Ilya, who'll go through the financial and operating results in more detail.
Hello, everybody. Let's start on Slide 4. In the third quarter of 2019, our assets grew by 12.3%, mostly because of the strong credit portfolio growth from RUB 290 billion to almost RUB 320 billion. There was also a contribution to this growth from cash balance, which grew almost 80% due to the SPO proceeds and strong inflow of current account money. As a result, the proportion of cash in the apps increased by up to 8% at the expense of the credit portfolio and debt securities portfolio. Other assets grew due to an increase of the security deposit in Mastercard. Moving to Slide 5. Our gross retail loan book added 10.8% in the third quarter and 79.6% year-on-year. This growth is the result of an increase in the credit card portfolio which gave us 480,000 new activated credit cards in the third quarter and solid growth in the home equity, car and point-of-sale loans. Cash flow in portfolio did not increase much in the third quarter despite strong loan disbursement. And the reason for this is the high early repayment rate for cash loans. The share of credit cards in the loan portfolio reduced to 62% of the net credit book. The net loan portfolio increased by 10.2% in the third quarter and by 61% -- 61.2% year-to-date, as Oliver mentioned. And we started giving more granular information on the loan sub portfolios that you will see later on Slide 12 and 15. The quality of our portfolio remains high, while the NPL ratio has gone up a little bit to 7.9%, and loans 90 to 180 days overdue going up to 3.2% of the gross portfolio. The NPL coverage remains in the 170% area. The group's funding structure can be seen on Slide 6. The total funding base increased by 6.1% in the third quarter and 49.1% year-on-year. Individual customer accounts increased by 9.6% quarter-on-quarter with both current accounts and deposits showing strong growth. SME customer accounts increased by 13.9% after first half slowdown as we continue to actively acquire SME customers. On the wholesale side of the business, we closed short-term repos and issued a RUB 10 billion 10-year bond with a 3-year put.On Slide 7, you can see shareholders' equity increased by 53% in the third quarter to RUB 83.4 billion, thanks to -- billion rubles, thanks to the SPO in July, strong quarterly profits and the suspension of dividends. Our Basel ratio has improved to over 20% as well as all statutory ratios. The N1.0 and N1.1 ratios went up to 11.9% and 8.4%, respectively. We have not fully utilized the incoming cash into the equity of the bank yet. When we do so in the coming months, it will bring our statutory ratios to 13% total capital adequacy and 10% base capital adequacy. On the next slide, you can see how the risk-weighted assets of the bank are distributed among credit, operational and market risks. During the year, our RWAs grew by devilish 66.6%, mostly because of the credit component, which has more than doubled. On the right-hand side of the slide, you can see the density risk weights in the retail portfolio and total credit market risk, which also includes other assets, cash and some securities. The trajectory of the retail book density will bring us to 175%, 180% in 2020. This is due to the gradual implementation of higher risk weights announced by the CBR, which will be, in part, offset by an increased proportion of secured loans with 100% risk weight. And now some comments on our profit and loss statement beginning on Slide 9. Compared to 9 months of 2018, our revenue grew by 46% to RUB 117.5 billion while the share of income from the loan portfolio remained almost flat in revenue composition at 68%. Credit revenue grew faster than non-credit revenue in the first half of the year. But in the second half, you can see that non-credit picked up pace. We expect this to continue in the fourth quarter as well. Going back to the credit business on Slide 10, you can see that interest income grew by 44% to RUB 79.1 billion in the comparable 9-month period. This growth corresponds to growth in the loan book and securities portfolio. Gross interest yield on the credit portfolio went down to 32.3% for the 9 months of the year, compared to 36.2% a year ago, mostly due to the growth of the non-credit card part of the loan book. And interest expense increased 40% year-on-year to RUB 15.5 billion. While cost of borrowing went down to 5.8% on a blended basis, as our funding from retail and SME becomes cheaper. On Slide 11, you can see that net interest income grew by 45% year-on-year to RUB 62.3 billion. Net interest margin went down to 22% and our risk-adjusted net interest margin decreased to 15.2% due to the reduction of the gross yield explained earlier and growing cost of risk, which went up to the 9% area during the last 2 quarters. Growth in the cost of risk relates to 2 factors: sudden worsening of risk across the market, and front-loading effect of IFRS 9. The latter contributes approximately 1/3 of the worsening effect. It's too early to give guidance about the direction of risk parameters in 2020, but we observed some decrease in cost of risk in October, November, back into the 8% area. The next 2 slides give more granular information about the unsecured and secured part of the loan book, including gross yield and cost of risk. Slide 12 shows the more mature unsecured loan book. You can see a mild deterioration, typical of having passed the middle of the credit cycle.And Slide 13 shows the younger secured part, where the risk parameters and NPLs are not yet fully shaped on fast-growing portfolios. Nevertheless, we can clearly see that this portfolio has much lower cost of risk and will have a positive impact on 2020 cost of risk dynamics. Our fee and commission income resumed its strong growth in third and -- second and third quarters, led by SME business and fees from current accounts. Wiring continues to contribute strongly. Fees increased by 34% in the comparative 9-month period and amounted to RUB 25.7 billion. The credit component of the fee income should be viewed together with the growing part of the other insurance premiums earned, where credit insurance contributes a significant part. Our current account business brings us more and more in terms of the number of customers and balances and transaction volumes as you can see on Slide 15. At the end of 2019, we stand at 6.3 million current account customers with almost RUB 170 billion of balances. The continuous growth of customers and balances allow us to build up a RUB 6 billion fee income, net of cash back that we returned to our customers. The current account business is breaking even taking into account post-margin contribution from the deployment of cash into the treasury portfolio and to the short-term portion of the credit portfolio. Our SME business is developing at a good pace, which can be seen on Slide 16, bringing us new customers, which, in turn, increase the transaction volumes. And at the end of 9 months, we had more than 500,000 customers with over RUB 45 billion in balances on SME current accounts. We earned RUB 6.7 billion in fees in the first 9 months of the year in addition to treasury income. The SME lending business is still in test mode with a total loan balance less than RUB 1 billion. On Slide 17, we give some operational statistics on Tinkoff Investments which had a step change in growth in the third quarter in transaction volumes and fee income. This business line is already bottom line positive, and will keep growing fast approaching 1 million customers very soon. Now turning to Slide 18 for some comments on operating expenses. In the recent quarters, you can observe a positive trend in our operating efficiency as our cost-to-income ratio went down from 41.3% to 38.4% for 9 months of the year. In the fourth quarter, operating costs might be elevated a bit due to the high season advertising activity and an annual salary increase in November of approximately 6%, 7%. Overall, we were able to show a quarterly profit of RUB 9.7 billion and RUB 25.1 billion for the 9 months of the year, which is 32% growth year-on-year. In 2018, 9-month growth in the bottom line was 31%, which proves our ability to post outstanding results despite significant investments into growth and increased loan loss provisions. Our return on equity went down from an abnormally high 80% plus into the 50%, 60% area due to the capital increase and to the suspension of dividend payments. At the same time, return on assets has reached a best-in-class level of 8%. And now back to Oliver.
Thanks, Ilya. I'd like to round up our session by refreshing guidance for 2019. We reiterate net income of over RUB 35 billion. Our net loan growth will be noticeably higher than 60%, cost of risk will be in the 8% area, and we expect cost of borrowing to come in at the lower end of the guidance in the 6% area. We plan to announce next year's guidance in March 2020, along with our full year results. However, I can say at this stage, that our 2019 business development efforts, coupled with a move into secured lending products, has laid the foundation for continued strong earnings growth momentum in 2020. Next year, we'll remain focused on our bottom line performance while continuing to build out the ecosystem. Our platform has become much more diversified, both in terms of our credit product offering and the ever growing suite of transactional and lifestyle services. As a result, next year, optimal earnings growth is now achievable even under a more moderate loan book growth scenario. This is especially important given the tighter regulatory capital requirements, which we expect to continue into 2020. Can also reiterate our long-range ambition to double the customer base to 20 million customers in the next 3 to 4 years. Thank you very much, and we're more than happy to take your questions.
[Operator Instructions]. Okay. So we will take our first question today from Andrey -- I beg your pardon from [indiscernible] Gazprombank.
I've got three questions, if I may. The first one is the deterioration of the quality of the loan book that you're observing. Is it's trajectory is better or worse than you expected? And how soon you expect this loan book to mature as regards to the risk crystallization? The second one is, how do you assess the capital ratio's effect from the new PTI regulation? And the third one is, are you planning to support your GDRs markets on the local -- on the Moscow Exchange? And if yes, how soon the liquidity could be better than now because for now, it was pretty weak, I think.
Thanks, Andre. So I'll take the first question, Ilya will take the second question, and Sergei will take the third. So one, I think your question was about not the duration of the loan book, but about the maturation of risks in the loan book. So obviously, we can't talk about homogenous loan book anymore. So we've got several different loan books, yes? So we'll try and break this down a little bit. In terms of unsecured loans, we have a very mature credit card business, which is being tested by two fairly deep crises and therefore, has been very seasoned. So there we're very good at, let's say, modeling risk, underwriting risk, managing the portfolio and collections, if required. So there, we're basically more or less where expected to be. All of the numbers tend to move around, as you know, for seasonal reasons, operational reasons depending on channel mix and whatever. This is something we repeat very often. So I won't bore you with all the gory details. But basically, the one number, which went a little bit soft is roll rates. So this is when a customer goes into delinquency, they become a little bit more difficult to collect, to return to current status. That's something that we're seeing across the market. We started seeing it basically at the beginning of this year. It's still continuing, but only, I would say, less so. So the vintages are stable. And if you look at what we've been doing over the last 6 months as a result of some of the tightening up of our underwriting that we did in the summer, some of it in late spring, you can see that, overall, we're holding our position there. So nothing untoward. The same applies to personal loans. But there it's obviously less of a seasoned business, and so we're keeping a very close eye as to how the risks develop. But again, if the vintages got a little bit worse at the beginning of this year in personal loans, now they're back down to where we want them to be. So we're basically, I can confidently say that vintages are stable at this stage. In unsecured -- the unsecured -- sorry, the secured loan book, so in car loans and home equity, there, the risks are developing because these are new books, yes? So they started at a very low level, they're moving up as the risks develop because these got a much longer, if you like, gestation period because of the types of loans that they are with the longer duration. But obviously, they're secure. So there we'll see how they pan out. The one thing that you have to understand about all of the business that we do on -- in all of the credit business lines, is that we apply a very high hurdle rate, this 30% hurdle rate in our NPV models that we've referred to sometimes, which gives us a very healthy buffer, should we go into a more difficult time in the cycle or more likely to be some kind of economic problems related to an external shock. And we've also been keeping well on top of underwriting an asset quality coming in. So you see our approval rates have been going down, especially in the unsecured lending products, but in other products as well as we make sure that we -- all of the incoming vintages are of comparable quality. So we're pretty happy with where we are. This is, as we were at pains to emphasize quite often, this is now the second phase of the cycle. So risks are going to go up. The question is with what speed? And we're pretty happy with what we're doing at the moment to make sure that we're well on top of it, so that risks don't deteriorate further. I don't know if Ilya wants to add anything to that and so on the first question, in that case, then this PTI and capital effect. Over to you.
Right. We are living this new PTI regulation for more than a month now. But honestly, it's still too early to give any quantitative statistics on how it affects us because it really should be viewed on the vintage basis. Give you just a simple example for that, so for example, we sometimes ask our customers for the form from their employer about their salary, and it could be this form and the [indiscernible] can be presented later when the loan is already issued by a customer. And therefore, even if, for example, at first moment when we issue a loan we have this customer falls over the PTI -- certain PTI threshold, then it comes back when a customer brings that salary form. So it's really too early to give quantitatively. Qualitatively, we are not much affected by this regulation before because it's always been in our scoring parameters. We've always been very cautious in giving loans, a significant amounts to those who are over-levered or whom we can make an over-levered customer. So nothing changed here. Of course, in terms of operational work, it requires certain adjustment to be made because we have to request different -- well, similar information from different sources and polishing this operational work, of course, takes time and happens as we speak. Of course, we are, for the time being, we'll be issuing certain portion of loans which we know are basically not levered. We know it from our models. But because of the absence of some information, they are coming out in the reporting as over levered. So -- and that's, obviously will have some negative effect on the growth of our RWAs. But as we noticed, these customers have a good credit quality and are NPV positive for us even with the higher capital requirements. We will continue to issue certain -- certain portion of these loans and our operational effort, of course, works towards minimizing this portion. So yes, we will be later on -- maybe it will take us a few months to basically to give more detailed statistics and how it all goes.
Just as I'm handing over to Sergei, so we're seeing a fairly weak effect on us in terms of additional risk weights. And certainly, having no impact on the approval side because of the fact that we baked this into our underwriting algorithms anyway and had done for many, many years. Sergei?
Yes. Just a quick comment on the Moscow listed GDRs. You're absolutely right. It's been only a dozen sessions on the Moscow Exchange, and we can't expect a major strong liquidity given the fact that this is a secondary listing. So there will be a natural stabilization period when a certain volume of GDRs migrate to a new center of gravity here in Russia. So in fact, we have been surprised that the liquidity has been improving quite constantly and quite rapidly. There's been a few sessions where liquidity achieved or daily trading volume achieved in Moscow exceeded that on the London Stock Exchange, which is good. We see a growing number of accounts and trades. So the positive trend is clear. So I think we'll have to be a little bit patient and see when this critical mass of locally domiciled liquidity forms itself. We've asked a couple of great brokers help us with market making. So frankly, liquidity should not be an issue for local investors preferring to buy Tinkoff through ruble-denominated Moscow listed paper. So I think we're quite happy with what we've seen over the past 2 weeks.
Our next question today comes from Elena Tsareva from BCS.
Congratulations with record profits. And my first question is about like the overall credit quality worsening you see now -- you observe. If you just can share any like quantitive indicators in the sector. What signal about like overall credit worsening? And if you may say, how are the peers, competitors react to this kind of worsening? This is my first question.
Yes, so how other players in the market responding is probably best a question you should ask of them. But generally, as a general observation, and based on conversations, obviously, we have with our colleagues in the market, roll rates have deteriorated somewhat across the market, particularly in the unsecured loan categories, which has meant that some banks have tightened up on approval rates, some banks have tightened up a little bit on the collections. And that's something that I would expect to continue. But this is a mild deterioration yes? So there's nothing moving south very quickly. Is it just something we should expect, given where we are just past the midpoint in the cycle. We do not believe that we're late stage -- second -- late stage in the cycle. Or people put it in various ways. But basically, I've read a couple of the reports that have come out today. We disagree with them. So the market is tightening up and obviously, there are a number of different reasons for this. One of them is the slight deterioration of asset quality, which is as a result of roll rates, which I believe is a function of increasing leverage in some pockets of customers. But it's also a result of risk weights, which have a cumulative lagged effect and are kicking in as a result of the full risk weight increases that have happened in the last couple of years, but also of PTI. So some banks will have responded if they don't have -- if they don't have the capital to deal with the additional risk weights on top of everything else as a result of the PTI regulation, which came in from the 1st of October, and/or they have a very high proportion of higher PTI, i.e., over the 50% threshold applicants coming in their application inflow, then they may be cutting back on approval rates, which will have a drag on growth. So I would expect to see the -- what we believe is already happening, a deceleration in the market as a result. In terms of what we see, and I'll just reiterate what we said many times over the last few months, our approval rates have come down, which is a leading indicator, and that's come down because the number of applicants in our application inflow who have a higher PTI or DTI or a number of recent inquiries or some other things which look like based on statistical scoring models that have a higher probability default. It means we won't approve them. So our approval rate is readjusted and gone down. So if the highs of 25%, 30% in credit cards, for example, a few years ago, we've gone down to basically around 17%, 18% approval right now. But it moves around, and it moves around for a load of operational reasons because of channel mix and all do -- all the good stuff that you know. Delinquency rate is going down in credit cards, for example, so the delinquency entry rate -- entry rates. So, i.e., the rate at which customers in a bank book go into delinquency is actually still somewhat improving. It's not deteriorating. But once a customer goes into delinquency, then they are somewhat harder to collect. We think we've stabilized this, but that was the number that was moving a little bit out of whack. So that's what we see. The vintages are stable, they're not deteriorating further at the current time.
Just maybe a second question, like on cost of risk dynamics, which we now see somewhat a bit like -- a bit stabilized. It's just increasing, but it did stabilize for the quarter. And like what you explained toward the -- about the market, especially given that like the mix of products on your book is changing towards more like secured. Maybe you can just show how you see maybe just how much maybe there is increase from this level of cost of risk or maybe we can see even more or less stable from the current level like at least for one year?
Yes, Elena. Well, it was a little bit difficult to hear sometimes what you're asking, but I...
I am very sorry. I am at the hospital with my kids. I'm very sorry, just..
Yes. It's about the cost of risk. So we see that right now, in the fourth quarter in just 45 days of the fourth quarter, we see that the risks -- cost of risk is improving, as I mentioned, into the 8% area. So that, at one hand, it's because, as Oliver mentioned, the risks are right now stable. On the other hand, there could be a seasonality because we know that fourth quarter is always a lower credit risk in the year. But it's -- we usually see it closer to year-end in December. So we'll see. We also know that the first quarter is usually worse cost of risk compared to the fourth quarter of the previous year. So it will be changing at that moment. We are not giving any guidance right now how we see the risks in 2020. Of course, I can only state the obvious that our secured loans have lower cost of risk than unsecured. And therefore, if they will be developing at a faster pace, they will basically have positive -- that we will have a positive effect on blended cost of risk. [indiscernible]
Our next question today comes from Andrew Keeley from Sberbank.
I've got a couple of questions. First of all, on your gross loan yields. As I remember from the second quarter, there were some kind of one-off factors that basically pushed the yield up a bit. And it looks like if we kind of take that out, I think the loan yields fell something like 180 bps or so Q-on-Q, which is still quite a bit more than perhaps than I was expecting. I'm just wondering if you can kind of give us any sense of how you see the relative kind of importance of changing mix effect and lower interest rates in terms of putting this downward pressure on the loan yield? And do you think that the kind of extent of the fall, even if we adjust it in the third quarter, we shouldn't really see such a strong fall kind of going forward? Just would be good to get some thoughts on that.
I'll start answering this question. So you are absolutely right, there was certain one-off effect in the second quarter. We did not have it in the third quarter. So yes, you can basically, in your modeling, you can take it out completely. You're absolutely right in saying that changing product mix has basically the dominant effect on our blended gross yield going down. And we see that in the sub portfolios, individual sub portfolios, yields are more or less stable. There was no specifically strong trajectory downwards. As, for example, it was the case a few years ago when our yield on credit card book was going down steadily. Right now, it's more or less stable. Though I'd say, of course, in the long run, taking sort of a few years view it should go down a notch further. But right now, it's all about the change in mix. So in the unsecured portion, for example, the point-of-sale loans have a lower yield. And therefore -- and we were very successful in distributing point-of-sale loans in the third quarter, and therefore, that has ceratin effect as well on unsecured.
So just to pick up on that point, and your right, so as Ilya says, this is a function of change in the loan mix. So we have some lower-yielding secured loans. We have low-yielding point-of-sale loans. When you look at the yield on the individual lending lines, they're not going down. And just to kind of step back from this now to give a little bit of context. Because as you know, we don't compete on price. So there are some segments in some lending business lines, which are slightly more price sensitive. But they have a let's say, less of a visible effect on the gross yield. And so generally speaking, we have not been reducing our rates in any of the product lines. In fact, some of them we've actually managed to increase them slightly. So what you did see a few years ago, if I recall, 2 years ago, maybe a bit more 2.5 years ago, was our response to, if I recall, the second increase in risk rates. Our response was to move down, for example, our credit card loans when they were originated to the lower PSK bucket. That's not something which we've done since. So we price adequately for risk. We're not entering a race in terms of pricing, especially given where we are having passed the mid point in the credit cycle. So this is not us, let's say, pricing down or our customer base behaving in a radically different way, this is just the lower mix.
That's helpful. I guess just coming back on that. I mean I think in the past, Ilya kind of talked about expecting 2% to 3% roughly in terms of yield decline annually. Now that may be changing because the mix is changing, perhaps in a way you couldn't tell or didn't expect. But I mean it looks like -- certainly, if we look at, say, the first quarter, third quarter, the decline has been quite a bit stronger. I mean I don't know whether you have any sense on whether you think that perhaps it should kind of ease a bit from here or this is the kind of pace that we should get used to?
Well, I would say the same right now, that the blended yield will be going down, same speed, as I mentioned, but they are more -- sort of more information that we now give on different subportfolios. I can give you an idea that the yield on secured part is not going to go down. It's going to remain in the same, it is right now. On the unsecured part, it will probably go down because of the, again, changing mix within the unsecured because there are different price strategies there. But it's definitely going to be lower than 3%, probably 1% a year, something like that. But again, it really depends on a few moving parts. So how fast are we growing credit card book, how fast are we growing point-of-sale lending, what's going on with the cash lending business. So it's again, that's a mix.
That's fine. And just the second question is on costs. Obviously, your customer acquisition costs fell very nicely in the third quarter. I think, Ilya, you mentioned that there will be a bit of a pickup in the fourth quarter from some kind of seasonal advertising and such like. Do you expect that, that will be a shorter effect this time around than we've seen over the last year, where basically you had quite a prolonged kind of spend uptick on advertising, it seems, and there were three, I think, pretty strong quarters in terms of the marketing costs? Is it the case that this is more just a kind of one-off end of year seasonal thing rather than a kind of another pickup in acquisition marketing costs for a longer time period?
Well, the marketing costs, we'll spend more in fourth quarter, and we will spend less than the first quarter as usual. So it has short effect. The increase in salaries, 6% to 7% of that, so 6.5% on average. It has a long rate there to the contrary simply for the fact that we're doing it in November. So it basically, it's effective half in the fourth quarter. And then it's full in the first quarter of next year. So it really takes a couple of quarters for the revenue growth to overcome this step increase in salaries. But otherwise, I'd agree with you that this time effect will be shorter and you, basically, you will see healthy numbers in terms of cost efficiency next year as well.
Again, I think it is important to step back from this, Andrew for a second. So you're talking about quarter-on-quarter, seasonal stuff and whatnot, which is important for you to understand the dynamics of what you're seeing. But to step back from this and think what's happened this year and what will be happening next year. So just so you understand some of the underlying trends that are going on here. So the first half, we spent loads on brand advertising, on TV as part of that as well, on customer acquisition, because we're in a very rapid growth mode, especially in the first half, but going into the third quarter as well. Next year, we'll be growing at a lower rate. We're not guiding exactly for how much yet, but it will be low. And as a result of the top line growth that you see this year continuing into next year, strong top line growth. Obviously, that dilutes what we're spending on customer acquisition. Brand expense will be maybe in absolute terms, a little bit lower than this year, but it certainly won't grow. Salaries will go up in the fourth quarter because we have a salary review in November, and bonuses are paid if we have a good year. So obviously, that part goes out and if you're looking quarter-on-quarter, but just to give you a bit of color on our salary growth for the year, we've actually -- we brought in loads of new people, especially into HQ, especially IT people on the tech side in basically 2017, 2018, it's very notable salary growth, as you recall. We slowed that down in the fourth quarter, basically Autumn of 2018. It's been a lot slower in 2019, and that will continue in 2020. So all of that has helped contribute to our reduced cost-to-income ratio. And that's [indiscernible] spend.
That's very helpful. And I guess, we can assume that you would be hoping, expecting the cost to income ratio will fall further through next year?
We're not guiding for cost to income ratio, but we might assume that at least this won't be growing.
[Operator Instructions]. We will take our next question from Mikhail Shlemov from VTB Capital.
I recon I have a couple of questions. The first, which I would like to start with is actually trying to understand a little bit of the dynamics of actually fairly strong growth, which you had in the credit card portfolio in the third quarter. Especially, I was wondering how the utilization rate on the credit cards has changed, perhaps either in the Q3? Or is it easier for you to get the number, how it has changed, basically the year-to-date? The second question refers to the non-credit lines of the business and specifically, insurance and retail investments business. I think on the retail investments you will think that the business has broken even in October. So perhaps you could guide us to how -- what -- to how meaningful contribution we could expect to the bottom line next year? And the second part of the same question is basically of insurance. You had or actually quite a big pickup in the premiums specifically other than the car insurance. I was wondering what exactly has changed in the business and new products, which you have started to offer? Or how you think about this business and whatever it would be also be a breakeven or even a bottom line contributor next year?
Thanks, Mikhail. So on credit card utilization, as you know, we look at the NPV based on a lot of different metrics. So in some channels, you get lower utilization, the lower cost of acquisition or low-cost of an application form, and maybe lower risk. In other channels, you may have higher utilization, but maybe higher risk. So it's all a combination of these factors, which we'll take into account. We calculate them. We understand them. How many customers we can get in different channels is a moving feast, obviously. On what their NPVs, we rank the different channels by NPV, decide where we want to, basically, if you like, invest our resource, and we bring those customers in. So utilization rates depend on different channels. But as I said, there was some lower utilization rate channels and some high utilization rate channels. All that said, our utilization rates in aggregate across the product, across credit cards for the year has stayed high. We haven't seen a drop in utilization rates, we haven't seen anything untoward happening and they're not going up or not going down. So basically -- so simply it's a metric that, obviously, we're very sensitive to because it's one of the things deeper into the marketing funnel, the sales funnel that determines the cost of acquisition of a customer. So we keep a very close eye on it, and it's now moving.Your question on investments, I'll hand over to Ilya in a second for that. In terms of insurance, we did quite a lot of testing in spring. So KASKO, the voluntary car insurance product has been growing. It's been growing from a higher base or certainly not a large -- a big base, but at a slightly lower rate maybe than last year, but we learned basically about a year ago, maybe a bit more now, how to book NPV customers in KASKO for voluntary insurance and car insurance. We've been scaling that up, you'll see more growth coming through in the second half of this year, it's something we like. We know how to do it. We know which channels to put customers in, we're experimenting with new channels, and this is going to be the main driver of growth in the car insurance business. We'll break even, basically, first half next year. And our expectation is that this should be a mild bottom line contributor on the non-captive insurance next year. All my finance colleagues say will back me up then, nobody disagrees with that? Okay, good. OSAGO, which is the mandatory car insurance piece, we did a large test in spring. We booked a load of OSAGO customers. The jury is still out as to how successful this test was. So we're still gathering and analyzing the data. We slowed that down a bit again, and we'll probably come back to the OSAGO market next year, maybe using a slightly different tack. But that also explains some of the increase in premiums that you saw -- that you're picking up on, Mikhail. So we like KASKO. We like the insurance business. I'm talking about the non-captive insurance business here, which is what we run the business to and the team -- we hold the team to account on those metrics, not on the captive insurance business, obviously. We -- as you know, we're firm believers in the longer-term prospects of this, which start to keep plugging away. Ilya on investments?
Right. Well, we broke even in July, not in October. So the third quarter was a positive bottom line contribution from Tinkoff Investments. Despite the fact that they are booking customers at the same pace as they did in previous quarters, more or less, so very successfully. We are adding a lot of new features that will allow us to earn more, for example, short sales, for example, our own of Tinkoff Capital, ETFs. We have a few cool features in our sleeve to basically to make these ETFs growing very good and profitably for us. Well, without giving -- we shouldn't give guidance at this point, but we are sort of -- mentally are programming our guys in this investment business that next year, they have to give us something about RUB 1 billion in operating income. But it's not a guidance. It's where we're basically, yes, where we went -- we want them to be in a year's time.
Okay. That's really helpful. I'm sorry, if I may also add another question actually, just like a follow-up to what Andrew has been asking about the growth yield. Perhaps if you could give us some color on how the interest rates on your front books, basically the current loan mix which you are issuing, let's say, for example, in the Q4 with a new regulation, compares with interest on the gross yields on your back book, which we have been seeing so far? Just like to give us a sense of...
Exactly...yes. Sorry, Mikhail, we understood the question. Exactly the same, no change.
So it's basically all the same on front and back book, right?
Yes. Exactly. So just -- we've been asked a couple of questions on this call. We've been asked a few questions recently. What is the impact of PTI regulation? Yes. So people think that we've slammed on the brakes. So it was one of the questions. The other question is, have we changed our gross yield and pricing. Other question is what do we see on the risk side, what's up in the approval rates, et cetera, et cetera. So the answer is that because we baked in PTI, DTI inquiries, number of loans and a lot of other parameters into our -- underwriting parameters into our scoring for many, many years, there has been no change to our approval rate, to our growth rates, to our pricing as a result of the introduction of PTI from the first of October. There will be a change, as Ilya mentioned earlier, to our risk rates because we are issuing some loans in the higher PTI category, i.e., above the 50% threshold. Maybe, over time, we'll do something different with those higher PTI loans, but right now, we're booking them as we were before because they're low-risk and high NPV, and we've just taken a hit on capital. That is the only change to what we're doing. Apart from some operational stuff around collecting confirmation of salary and that kind of stuff.
Our next question comes from Andrey Pavlov-Rusinov from Goldman Sachs.
Congratulations on the results. Most of my questions have been answered, but I've got still a couple. First of all, on the growth -- on loan growth, I kind of understand that you will be giving next year guidance in next March. But essentially, maybe just some early thoughts on how significantly can growth slow down next year? For example, the issuing rates that we have seen in the third quarter, there was still a slowdown relative to the first half of this year. Could this be indicative of what we should expect annualize next year or the growth will be significantly slower in -- especially in the, say, more of our established businesses like credit cards? My second question is more specific about cash loans, essentially, there was a very strong slowdown in issuance and the cash loans that you made in growth of the book. What was the reason for that? And also why if the growth has slowed down a lot, the cost of risk in this particular business line didn't really drop at least because of lower IFRS 9 provisioning? And my final question is on insurance. You've covered a lot about the auto insurance. But what about the credit protection insurance. It looks like it increased a lot by at least RUB 1 billion in the gross premiums throughout the third quarter. So any color on what's happening there would be very helpful.
Right, Andrey, I'll start with the loan growth. And you're absolutely right. We are not in a position to give a specific guidance. So I just -- will be talking in more general terms. And basically, I'll start with the credit card book. And of course, we have change in channel mix from year-to-year, but we are able to show pretty good growth in credit card book. And I don't see any specific reason for us to basically to stop increasing our credit card loan book by same -- basically, you have to have similar expectations about our growth of 20% to 30% as we were able to grow in the few previous years on average. As for the -- our new products, home equity and car loans, here we will, of course, we'll be growing much faster because of the lower starting base. And here, I'd say we don't see any obstacles to show the similar disbursements as we did this year. So basically, we will -- the issuance would be at least the same, hopefully, but what it would mean for the portfolio sale at the end of 2020. If the issuance is the same, then the portfolio growth is not 100% because of the repayments, right? So it will be growing faster than unsecured, but to what extent, it's difficult to say. But the issuance could be absolutely similar. We see the channels. We see the economic. We see the risks are very good, the acquisition cost is stable. So very, very good here.On point-of-sale loans, we are working in this segment with very good results. We are amongst, probably top 3 players in this market. It's an acquisition channel for us. So we can either increase the -- issuance have slowed down a little bit, take some technical moves, but again the portfolio is very fast turnaround. So therefore, you shouldn't probably expect very significant growth in terms of balance here. But it brings -- so this business is about the number of customers, not about the balances. So in my view, the balance here is not important. And if it's not growing, we will have no problems with that at all. In terms of cash loans, that's probably the most difficult to predict. It might grow or might stay in a similar amount. And that would be basically the same what happened in the third quarter of this year because we had a strong disbursement, a bit less than in the first half of the year or first, second quarter, but still very strong disbursement, but very high early prepayment on these cash loans. So we issue them 80% of these loans, are issued for 3 years, 10% for 2 years, 10% for 1 year. So that's a initial statistics. But when it comes to a 2-year horizon, basically almost all the vintages are repaid. So there is basically a remainder -- a tale of 10% or something like that. So the prepayment -- early prepayment rate is very high. And that's a serious constraint to the growth of the balance -- annual balance of this book, which is from one hand, if it allows us to make a positive, healthy NPV, and we see our customers repaying our loans fast, then it's basically a good thing. So we -- again, we're not worrying about this. And for us, making this somehow grow artificially by either increasing size of the loans or increasing the tenor is not what we're going to do.And from this, I'll step into the second part of your question about the third quarter, and I'd say that, of course, there is a factor of worsening increase in risk environment because, of course, we've changed our underwriting parameters a little bit and reduced approval rates, which had an effect on a slowdown in the cash loan portfolio at one hand. But still, I'd say that the disbursements were very strong anyway. So we -- so most of the slowdown is a factor that the cash loan portfolio was growing very fast in the second half of last year and the first half of this year. And therefore, basically, the early prepayments basically caught us up like a wave at some point.The third part of your question is about the credit protection insurance, which true is -- basically is increasing on the insurance side of our business. So on Slide 14, you can see that this -- the commissions that bank is getting. So basically, they create protection, which is sitting on the bank side, is not particularly growing in the last few quarters, but the insurance side is growing faster. So that's in general, should be viewed as a whole. It's basically 2 sides of 1 coin. But we have a certain pickup in the insurance premiums coming from our secured lending because it's -- well, it's a healthy part of the product proposition in the home equity lending for these reasons.
That's very helpful. So just to confirm, basically the -- as for the cost of risk, not dropping because of the IFRS 9. So it's kind of logically the faster rate of repayments also, should be positive for your models for the IFRS 9, given that actually the loans are lower duration, in fact. Should we expect some positive effect next year from that or not really?
Well, faster early prepayments is not necessarily improving our cost of risk. To think about this mathematically, the good customers are paying faster, but the customers who went delinquent, it has nothing to do with them, unfortunately. So it might have absolutely opposite effect. But it's -- the early prepayments right now, the effect of this is negligibly small on cost of risk, I think. So yes, you're right that the fast growth in the previous few quarters had a serious effect on front-loading some of the reserves due to IFRS 9. That's true. I guess if we would -- it's a little bit difficult to predict here. But again, if the disbursements are high and the prepayments are high. And then again, they share a fresh sort of front book in total, cash loan book is higher, and therefore, the IFRS 9 effect will be there longer. But it's -- we have to work on this to basically to calculate the exact part here.
[Operator Instructions]. We will now take our next question from Andrey Mikhailov from Sova Capital.
My question is on dividends. While you don't give specific projections for loan growth rates, it seems that they will be coming down. And this is probably the reason why you plan to restart paying dividends at a 30% payout ratio. But my question is on more distant future. Do you see loan growth rates falling enough to start paying the dividend at a 50% payout ratio, maybe somewhere towards the end of 2020 or early 2021?
Andrey, as discussed on a few occasions, our dividend policy is a function of quite a few factors. Basically, our key priority has always been to make sure that our growth is adequately capitalized, and we're talking about growth on the credit side of our business and noncredit initiatives. So now that we're in the second phase of the cycle, of course, the capital buffers consideration for a hypothetical deterioration in cost of risk is yet another critical consideration. So whatever residual capital we have available gets distributed to the shareholders. So this philosophy has been with us for many years now, and we'll put it on hold for couple of quarters, as you know, only around our SPO. So the dividend have been put on freeze for the quarter.The 30% payout ratio is the current level that we think the sweet spot for the next few quarters, where we'll find a good balance between capital allocation for growth, capital buffers, as I said, and also, we have to assume further tightening of the regulatory environment, which also it's up your capital position. So 30% right now looks like the sweet spot. And we are likely -- we, as a management team, are likely to be recommending to the Board to consider the renewal of or the dividend payments next year based on the Q4 results. Beyond 2020, it's too early to say quite a few moving parts there. And I think this is a subject of a nice discussion 12 months from now.
Thank you. That will conclude today's question-and-answer session. So I'll hand the conference back over to your host for any closing remarks. Thank you.Okay. It seems today's conference call has now ended. Thank you for your participation, ladies and gentlemen. You may now disconnect.