All right. Thank you and a warm welcome to all of you to this Q4 presentation. As always, thanks for spending your time with us today. We find that the Q4 release is an excellent time not only to cover the recent performance, but also to summarize what happened in 2019. This time around, we would also like to take this opportunity to share our view on the outlook for 2020 and what you can expect from Hoist going forward. On top of our financial review, we will consequently also share with you today our financial targets for the coming years. Kristy will cover that in a minute. I'm here today in Stockholm with Kristy Watson, as usual, and Andreas Aspgren, our Investor Relations. Moving on to slide number four, which summarizes in numbers the Q4, but also the full-year performance.
I think we can start by saying that 2019 was a year where we decisively have been dealing with regulatory changes. We are consequently happy to report consistent and strong improvements across the board. I think we can label this as a year with a strong execution. First of all, since the second quarter of 2018, and despite the additional costs from new regulations, the performance measured as profit before tax is on a quarterly basis, consistently around SEK 200 million. This shows that executing on our strategy works. Secondly, we are very pleased with the growth in 2019. In a year where we have been heavily impacted by regulatory changes, we are still able to grow the book value with 18%. Actually, our fourth quarter was the biggest quarter ever for the company as far as portfolio investments is concerned.
Q4 portfolio investments made up for a bit more than 50% of the total investment for the year. Let me also add point number three here, and that is that our share of digital collection continues to develop favorably, and more about this in a minute. Moving on to slide number five, as I said, 2019 has been a year of execution. The team has delivered, I would say, basically on all fronts. We have progressed in a number of key areas to deliver on our strategy. Let me just call out a few items on the different cornerstones of our strategy. First, regarding market leadership, we have said clearly that we prioritize being concentrated and not scattered, and that we are aiming for the top three positions in our core markets.
The GetBack transactions in the spring, and of course, the large secured NPL transaction that we did in France just in November, makes us take important steps in these two core markets for us. We now have the clear number two position in Poland, and we are actually the market leader in secured non-performing loans in the important French market. I think in this industry, we just have to agree that having the lowest cost of funding and having access to funding offers a key competitive advantage. Of course, having completed now our first rated securitization in the fourth quarter is a significant achievement and proves that our banking platform continues to offer unique competitive advantages. As a point number three, the digital transformation in the industry is happening now.
We are the frontrunners in finding better ways to help our customers keep their commitments. Our ambition here is to be the digital leader in our industry. Have in mind that the digital is more flexible, more interactive, we learn faster, we deploy quicker, and of course, to radically lower cost compared to the old ways of conducting our business. As a point number three, I think it's important to be more effective, more efficient, and again, the winner in the industry over time will be the ones who have the best operations. I'm happy to say, actually, that this work is progressing as it should. Kristy will revert to the numbers and share more with you. Let me assure you that our actions have been taken, but most of the benefits come in 2020 and onwards. More about this later.
Moving to page six. For us, consistency matters. Hoist has been growing year-over-year, and the progress is evidenced by these two graphs. Adjusted for items affecting comparability, but not for the impact of regulatory changes, the charts clearly show the improvements. Compounded annual growth of 16% for adjusted EBITDA, and for profit before tax at 25%. It's, of course, great numbers. It is significant and no coincidence. I would argue that this shows our commitment, our resolve, but also the resilience of our business model over time. Slide seven is a brief case study of what's been happening in France for Hoist Finance. As you can clearly see from the chart, the book value has grown about five times since 2017, and our strategy of growth into adjacent asset classes is proving to be right.
In 2017, we start building the team in secured non-performing loans. In 2018, we bought our first portfolio. Now in 2019, we have become the market leader in the secured non-performing loan collections. Actually, it's really fun to think about the fact that France now is our fourth biggest market. France remains one of the most interesting growth markets in Europe. Slide number eight. We have talked about our strategy of being the digital leader in our industry. Everything we do starts with our customers. By increasing flexibility and tailoring solutions to the customer's needs, we help people back to financial inclusion faster and in a sustainable way. Before Christmas, we got a new employee. He's always polite, he's keen to help, learns fast, and never gets tired. His name is Kai, and he is a chatbot.
If you look at the screens on the slide carefully, you can see my name there. What you see here is actually my user. It's a test user, of course. I don't actually have a debt to Hoist or to Robinson Way. You can see that we are actually trying to understand how our digital channels actually work. We're going to share more about this in a later presentation to tell you more about this. We think this is quite unique. The interesting thing is that Kai, as a chatbot, now actually handles more chats than our human agents. Actually, more than 60% of all chats are now handled by Kai. This, of course, frees up agents' time to deal with more complex situations.
Again, we would like to spend our time in the best possible way to serve our customers and to help them navigate through difficult waters. Moving on then to slide number nine. A year ago, we outlined our plan to mitigate the negative consequences from the regulatory changes that we were experiencing. We are pleased to see that what we planned is actually executed and is working. Securitization is obviously our most important tool, but we also see very good progress in our work to implement more sophisticated internal risk models, IRB. As you can see, our target now is to complete our IRB application within 12 to 18 months. I already mentioned our efforts in secured non-performing loans, and Christer will talk more to the composition of the book in just a few minutes. Let's then move on.
Christer, I hand over to you to take us through the numbers.
Thank you, and good morning. Starting on page 11. Looking at the P&L for the quarter and the year, the portfolio growth comes through. Comparing 2019 to 2018, net interest income grew by 18%, which matches the growth in our loan portfolio during 2019. Collection performance in the quarter and the year came in at 103%. This is somewhat lower than last year, and as previously commented upon, it is primarily in Spain where we continue to see unsatisfactory collection performance. This also influences total operating income, which grew by 10%. Comparing 2019 to 2018 again, the expense level increased by 8%. That is excluding items affecting comparability. Since 2019 has been quite a busy year, we've had some of those items, as highlighted on the next page 12.
Here you have a summary of the reported figures all in. I will not dwell on 2018, just as a reminder for 2019, some of the larger Items Affecting Comparability was transaction costs in connection with the securitization. We also saw additional costs in connection with outsourcing of IT, which affected some 60 colleagues of ours. We've also taken costs for restructuring in France, that affected around 30 employees. I want to stress that these restructuring costs, which we've taken in 2019, they have not generated any benefits in 2019. We will certainly see those benefits as we move into 2020. In fact, already during January, the responsibility for IT has been handed over to LTI, we have turned off the lights at our site in Bayonne, in France.
Taking a step back, these are just two examples of a much broader effort which will drive cost income down. On page 13, we have illustrated the cost income ratio for 2018 and 2019, adjusted for the one-off items I just mentioned. As we have clearly stated, we find cost income of 73% to be way too high, and we have also shared updates on the good progress that we are making on actual costs. Still, 2019 cost income level is on par with 2018, and this apparent contradiction relates to the fact that in the bank statement, funding cost is part of income, and the more expensive funding in 2019 has, in that sense, reduced income, the denominator. Without the more expensive funding, 2019 would come out with a cost income of 71%.
Adjustments aside, our day-to-day focus is, of course, on what we can do to improve this, and we can do a lot. I wish to bring out three examples. First, on personnel expense. This is something we address by increasing the share of collections through digital channels. We've spent significant resources on this in 2019, and we can see the benefits growing month by month. Secondly, collection costs. These are external costs, could, for example, be court fees. In this category, our focus is to make sure that every penny is spent where it makes the biggest difference. When we put our vast pool of data to full use, there is room for us to be a lot smarter than we are today. Finally, administrative expenses. We've made good progress on this front by streamlining both our organization and our site setup.
We've not reached the end of this road, and we are immature, for example, in areas such as procurement. To sum up, we have no doubt that the cost income of 65% is achievable also with the current, more expensive funding. I say this knowing that several of the efforts that we have completed in 2019 have yet to contribute in run rate benefits. Turning to page 14. I mentioned court fees on the previous page, here I just wanted to highlight that they can actually vary quite a bit from quarter to quarter. These are costs which are expensed as incurred, they generate substantial future value. In Q4, it happened to be the case that we took extensive efforts both in Poland and in Spain, this additional spend level should not be understood as a new normal.
With that minor clarification, we wish to leave the topic of costs and instead spend a few minutes on our 2019 acquisitions, current composition of our book, as Klaus-Anders Nysteen mentioned, and the nature of projected cash flows. These are certainly areas where we strive for a high degree of transparency. Starting on page 15, our 2019 portfolio acquisitions, they added up to almost SEK 6 billion , and at this level, we grow our core business at a healthy rate. In fact, the investment replacement rate is around SEK 3.5 billion . It is particularly pleasing to see a healthy mix of unsecured and secured assets and an attractive distribution across our core markets. Both France, Poland, U.K., and Italy all contributing substantial investment volumes.
In relation to so-called forward flow contracts, one can note that as market conditions have improved, some of these old contracts have been renegotiated and some have been canceled. Continuing on page 16, we come into 2020 with a total NPL book accounted for at close to SEK 23 billion. This value is derived directly from the estimated remaining collections, ERC, which per year end amounted to close to SEK 39 billion. The graph illustrates the distribution over time of these remaining collections and to exemplify if we were to acquire no new portfolios in 2020, we would still expect to collect SEK 6.5 billion in 2020 on the portfolios that we already own. That corresponds to SEK 3.2 billion of income already in the bag. This is, of course, a significant amount, and it's one which is underpinned by millions of customers.
Each month, we interact with roughly 170,000 of them, and over a year, we set up around 200,000 payment plans. The typical payment plan for an unsecured debt is around EUR 54 a month. When we aggregate and translate this into cash projections, we do so based on more than 20 years of experience in pricing, acquiring, and collecting on thousands of portfolios. Before we leave this topic of stable and visible earnings, a few final comments on page 17. Because even with 20-plus years of experience, your predictions will not always be spot on. For example, in 2019, we collected 3% more than predicted. We are continuously reviewing the projections that we have and adjusting if needed. As shown on the left-hand side, such adjustments have historically been limited when put in relation to the value of the NPL book.
As a short comment on 2019, I think we've already in Q3 mentioned that Spain is part of the increase in relation to previous years. In the appendix, we have also added additional disclosures on multiples per vintage. I won't go into that now. Turning to capital and liquidity starting on page 19. As seen on the right-hand side, liquidity has been partly normalized. Q4 was an active quarter in acquisitions, and we also saw net reduction as a result of changes in the securitization positions. Capitalization rates on the left-hand side were affected by these large acquisitions. We leave the year in the target range for CET1, although towards the lower end, and that's okay. Q1 is normally a slow quarter.
Needless to say, we steer the business with these ratios in mind, and that is true for CET1, but it's also true for the other ratios where we have the possibility to issue hybrid instruments. This is something that we assess in our continuous capital management, and we are currently about to engage with investors with a view to issue new subordinated additional Tier 1 notes during Q1. Continuing with funding on page 20. Unlike Q3, we have in Q4 had limited net flows on deposits. In Germany, we have lowered our offered rates since we were actually attracting more deposits than desired. The current offered rate on German overnight deposits is 0.25%. The mix in deposit duration has been kept largely unchanged in line with our preference. Together with the securitization, this meant that interest expense related to book value remained flat at around 2.5% to book value.
Moving on to page 22. As you heard, 2019 has been a busy year, and it's been a year of reduced regulatory uncertainty. We feel now is a relevant point in time to return to the topic of financial targets. When it comes to CET1 target range and cost income, we confirm our existing targets for CET1, nothing has changed. For cost income, I already elaborated on 65% being achievable despite the more costly funding. When it comes to return on equity and earnings per share growth, we have concluded to set ourselves a somewhat higher bar. We can see that the banking model continues to be attractive, something we can build on. We can also see our market developing in a favorable way with supply being healthy and margins improving. As demonstrated in 2019, we are well-positioned to capture such growth opportunities.
When combined with efficiency improvement measures, we set out to deliver ROE exceeding 15% and EPS growth of 15%. The definition of these targets are unchanged, and you will find the details in the footnote. In short, ROE refers to an ongoing basis, whereas EPS refers to compound average growth rate between 2018 and 2021. With updated and slightly more growth-oriented financial targets, it's relevant to also revisit our dividend policy. Here, we retain the long-term target to pay dividends corresponding to 25% to 30% of net profits. That said, we are now a bit more explicit on the considerations which will guide the board as they give their recommendations to the AGM. As previously communicated, these considerations have led the board to recommend to the 2020 AGM not to pay dividends for the financial year of 2019.
Based on the outlook for 2020, the board has concluded to extend this assessment to also apply for the financial year of 2020. We're of course happy to take questions on this, but first I will hand over to Klaus-Anders for a wrap-up.
Thank you, Christer. On page number 24 is the outlook and the key takeaways from today. As Christer was alluding to, we see a very healthy market out there. This is driven by regulatory changes in the banks and our sellers of non-performing loans to an even larger extent than before. We see a lot of great growth opportunities across our core markets and also across asset classes. We have strong belief in our banking platform. It offers us competitive and sustainable advantages with access to low-cost funding. The digital transformation is ongoing, and we are on track to becoming the digital leader in the industry. We also are quite confident in our cost savings initiatives, and we are on track to deliver on the 65% cost-to-income target that Christer just mentioned. With that, we open up for questions.
Thank you. If you would like to ask a question, please press zero one on your telephone keypad. If you wish to withdraw your question, you may do so by pressing zero one to cancel. That is zero one if you would like to ask a question. Our first question is from Borja Ramirez from Citi. Please go ahead. Your line is open.
Thank you. Good morning. This is Borja Ramirez from Citi. Thank you for your time, and well done on the results and the increased financial targets. Two quick questions from my side. Firstly, as per page nine of your presentation, I see you have increased focus on the mitigating actions, which is great. I saw that as per the presentation on the IRB, you aim to complete application in 12-18 months. I would like to ask if it's possible to provide details on the potential capital benefit and timing on the IRB. Secondly, as per page seven of your presentation, it mentions you could consider new markets in 2020. Could you provide further details? Thank you.
Well, thanks. I can see that page number seven potentially could have been written clearer. When we say new markets, we actually mean existing Hoist markets, but for NPL secured. How can we move NPL secured collection into new Hoist markets? I think that could be a bit clearer when we're looking in at hindsight, but we are just looking at how can we take our experience in secured NPL into the other Hoist markets. Currently, we are in Italy and in France, to some degree in the U.K. and Poland, but not to a large extent. To grow those markets is, of course, then a high priority. Your second question was around IRB. I think it's a little bit too early to go all out and promise too much. What we can see is that we have ramped up the project considerably within Hoist.
The data that we have is very strong, and they go back, I guess, 30 years or so. We have a lot of details, much more than any other banks that are looking towards the NPL space. We feel quite optimistic about the prospects of using IRB as a method to bring down equity and hence provide capital release. We will update you as soon as there's more specific, more concrete things to share
Understood. Thank you very much.
Our next question is from Rickard Henze from Nordea. Please go ahead. Your line is open.
Yeah, good morning. I was on mute. Just a couple of very basic questions. When you talk about 15% EPS growth from 2018 to 2021, can you just highlight how much of Items Affecting Comparability you have in the base and what EPS 2021 that corresponds to in your target?
The definition that we've chosen for this is excluding IACs. That's both for 2018 and 2021. In our cost savings points, we've outlined roughly what kind of cost to achieve we see. I think you actually have that in the appendix as well. I would say that the majority of that should be, well, partially been in 2019. The majority should be in 2020, I don't see very much of that ending up affecting 2021. In terms of EPS amount, I don't have the number in front of me now. Maybe Andreas can run the numbers in the background here while we take other questions.
Okay. My second one is on the CET1 ratio, which is now just above the minimum part of the range. How do you see that in terms of growth potential and acquiring new portfolios in 2020? Can you give us an update on your next step in terms of mitigating actions, and if you're planning a new SPV or anything like that?
Yeah. In terms of acquisition capacity, we did in 2018 acquire for SEK 8 billion, and in 2019 for SEK 6 billion. I say looking forward, we should be within that range. That should be doable also given our starting point in terms of capitalization. We do not have any sort of securitization transactions that are very close to execution. I don't expect to see those transactions in Q1 and probably not in Q2 either.
Potentially by the end of 2020.
Yes.
Thank you.
You're welcome.
Just as a reminder, if you do wish to ask a question, please press zero one on your telephone keypad now. Our next question is from Jamil Khoury from SEB. Please go ahead. Your line is open.
Thank you very much, operator. Morning, everyone. Just starting off with a high level question. What has changed in terms of your internal business models now with this updated three-year financial target? Is it more on the cost side? Is it on the income side? Just your reasoning around it would be very appreciated.
I think what we disclosed today in terms of new financial targets, it's no surprise to us. I think it's clear, and we have communicated that also in the past, that the margins are improving favorably for our NPLs. What we are underwriting now is better than the back book, which we think is of course a very healthy sign for the industry. That of course helps at the revenue side. Secondly, our efforts to become more effective, more efficient, they are working. I bet it's hard to see it in the numbers, right? It looks like we're kind of fixed on the same level in terms of cost income. We know for sure that this is going to help. A lot of initiatives taken in 2019 will bring costs down in 2020.
Of course, just a growing book in itself helps to bring out the benefits of scale. We feel confident that the combination of better margins and lower costs becoming more effective, more efficient and more digital is going to help provide and deliver the numbers that we need.
Thank you. Then, just on the back of what you said on the favorable pricing development. I'm not sure you've been talking about before has been the effective interest rate or the gross IRR here. In Q3, it's almost 100 basis points down sequentially, and now we're up 10 basis points. You haven't really absorbed the effect of the French secured portfolio. Where should that trend now, given mix, given pricing environment, et cetera?
Yeah. We give a few different margin metrics in our reporting. One of them is net interest margin, and that will of course, be affected by the funding cost as well. That's just worthwhile to remind yourself of. When it comes to the so-called effective interest rate, this is a gross margin on the book. What we've said in Q3 was that if we compare 2019 to 2018, on a like-for-like basis, we've seen margins come up by 50 basis points. That is our assessment for where we are now in 2019 versus 2018. Where this will go in 2020, I guess we'll see. I'd say that we see a very good pipeline, and that's a promising start.
Even though, if I can add one thing here, we've been debating and discussing with you quite a lot, regulatory changes and consequences for Hoist. It's not like the rest of the industry do not have regulatory challenges, right? Many of our competitors have felt the heat from regulatory changes as well. On top of this, some of them are quite leveraged and struggle to find a way to finance growth. Which means, again, that the competition that we experience in our core markets is quite rational and quite healthy. That supports our thinking and our thesis that the margin development is going to be quite positive, at least for the next few quarters.
Thank you. Just a follow-up, and also tying into my third question really on, obviously, it's impossible for you to say how the split between secured and unsecured will look like in 2020. Sort of rough estimate, and also, again, following up on the former question, what's really the gross IRR on the secured side, and how should that blend feed through to the group gross IRR? Is it possible to comment on that?
I can start at least with the mix. What we have seen now two years in a row is that we are doing almost, let's say, 60/40 in terms of acquisitions between unsecured consumer and secured. I think that's something we can replicate. Maybe we even trend more towards 50/50. We'll see. It depends a little bit about the opportunities out there, and we are reasonably agnostic. We have the same return on equity requirements on these assets. In 2019, we didn't acquire any performing loans. We're looking at a few situations where we continue to look at those into 2020. To have a blend between unsecured consumer, secured, and a bit of performing is really what we are looking for. Any comments on the margins there, Christer?
Yes. I think what you can say is that, yes, it's true that gross margins on secured assets are somewhat lower, but these assets are also less costly to collect on. You would expect to see a marginal cost income to be lower on those transactions. Actually, from a profitability perspective for us, these two asset classes are similar. In that sense, you shouldn't expect to see the mix in portfolio acquisitions to drive a very different development on the bottom line.
Is it possible to see anything on the top line now?
I think that depends a little bit on which transactions we actually do in 2020, and it's difficult to predict exactly how that could play out.
There is, I guess, a history now for two years where we have been buying secured. If history can tell you something about the future.
Thank you. That's noted. Just one final question, and sorry if I missed anything here, but you said something about issuing an AT1 here in Q1. On the back of that, just on the interest expenses here, we've seen a change in the deposit mix really throughout 2019, which has increased interest expenses. Could you just elaborate on, obviously, again, quite difficult to say, but how AT1 could affect the interest expenses going into 2020?
Yes. No, it is correct that we are about to engage with investors to issue this AT1 instrument. I think you will actually find something on Bloomberg out this morning on that topic. It's very current. One should remember here that AT1 instruments in our reports are reported as equity instruments. It means that the cost for AT1 instruments are not included in the interest expense line. They are deducted from retained earnings from equity. In that sense, you wouldn't expect the new AT1 issue to change the reported interest expense. When it comes to the deposit mix, we've had quite a change in that mix during 2019, and that has been a result of us concluding that we wanted to reduce the so-called interest rate risk in the banking book. That is why we have steered in the direction of longer deposit durations.
We're actually quite happy where we are now, so we don't intend to change this mix any further. That also means that if you look at the Q4 cost, that is reflecting the new normal, and I don't see that moving in any particular direction throughout 2020.
Thank you very much for taking my questions.
Thank you.
Just as a final reminder, if you do wish to ask a question, please press zero one on your telephone keypad now. Our next question is from Victor Hellman from Nordea Credit Research. Please go ahead. Your line is open.
Hi, thank you. Rickard Hellman here. I have a question regarding the cash flow. You mentioned replacement rate about SEK 3.4 billion. Looking at the cash flow for 2019 from operating activities is around SEK 3.1 billion, hence a couple of SEK millions in gap between there. Is this something that is important for you? Is it something that we could expect to close this gap?
The first question, is it important? Yes. It's important enough for us to have included a slide in the appendix on this topic, which is slide 27. I don't know if we can bring that up on the screen there. If not, we will find it afterwards. On this page, we've done an additional disclosure on the cash flows and how one should think about the cash flows in relation to the replacement rate. As the headline here discloses, there is significant excess cash generation. There is enough cash generation to replace the book, pay dividends, and grow a bit. Maybe have a look at page 27, and if there are any other questions on that, we're happy to help you afterwards.
Those are adjusted for non-recurrent costs or?
Yes. Correct.
Which you would then expect to be significantly lower in 2020?
Yes.
Okay. Fair enough. Thank you.
Thank you.
Thank you.
As there are no further questions, I will hand the word back to the speakers for any final comments.
Well, thanks again for participating on this call. If there are any more further questions, please don't hesitate to reach out. With that, I thank you for being on the call with us this morning, and I wish you all a perfect day. Bye-bye.