Welcome everyone to our 2018 results announcement. As you probably noticed as you came in, we're sitting surrounded by a major demolition site. After 40 years in Ballsbridge, we relocate to more suitable space, and Facebook get to move in their European HQ. For those watching over the webcast, that explains any extraneous noise. If there are thumps or bangs, it's the demolition folk next door. This is our last results announcement from this particular stage, and the end of an era as we transition the executive leadership of the bank. The results we announce today are strong ones. The business has performed well in 2018, and we are pleased to be able to increase the dividend by over 40%. Our capital ratios are strong, and the quality of the balance sheet improved.
This is a very good position that Bernard and Mark are handing over to their successors, who are determined to build further on this success. Mark left last night, and Bernard leaves when the regulatory processes are concluded. Colin is here as CEO Designate, and Donal as CFO, because you will see in the accounts, at the time we signed the accounts last night, Donal was Deputy CFO, we had the email in from Frankfurt, and he is now CFO, congratulations Donal on that new position that you have grown into. I should say that this is a very smooth transition other than for those of us who have to wear ties this morning in respect to Colin. There's a whole pile of new ties appeared.
We are delighted to have a management team of such caliber coming from within our own ranks. This morning, Bernard will open, Donal will take you through the financial performance, Colin will wrap up. Bernard will chair the Q&A, please be aware that Colin is limited in what he can say until the regulatory assessments are concluded. Bernard's done a superb job here over almost a decade. He was brave back then to run into a burning building. This was not a great place almost a decade ago. The bank is now, as you see, fully repaired, and better than it ever was. There are Davy staff here today. We had extra seats provided. There was such a large number wanted to come and see their new boss.
I would just say to them that a number of AIB colleagues have mentioned to me that Bernard has been the best boss that they have ever had, and you are lucky indeed to gain him as part of your team. Bernard, for the last time.
Thank you, Richard. Thank you very much. Thank you for the kind words. What I'm going to do is bring you through the highlights, as you said, for the last time. The good news is it's pretty strong performance from the bank. Hopefully what we'll be able to do is to demonstrate that along all the key lines and in all the key metrics that we judge ourselves, the strategy is delivering. The performance, as evidenced by 2018 results, confirms that we're well on track towards those medium-term targets. As Richard has mentioned, one of the key things for us was to make sure that the recurring profitability of the business would be demonstrated from 2014 on, when we returned to profitability, and I think this set of results does that. Profit before tax is EUR 1.25 billion.
We are still getting some ongoing benefit from the reduction in NPEs, there's some write-backs that arise from that directly and indirectly. The number is slightly elevated in overall terms, so it's a slightly higher level of profitability than normal, but the underlying business is performing very strongly. The loan book has grown on both a gross and net basis, and a performing and a net basis, every aspect of the book has begun to grow at this point in time, which is important. The asset quality, both in terms of the NPE reduction that's taken place, and in terms of credit quality of what's been written at the moment, is very strong. The book gets better and better as we grow it. The whole series of high-level characteristics from the performance are good.
This has allowed us to get to the point of sort of achieving a normalized dividend level. There were three key aspects to dividends that I think people were interested in. Firstly, was turning the dividend back on. Obviously, this is the third annual dividend, so that happened sort of three years ago. That was very important when it happened. The second point was getting to the stage where, on an annualized basis, the dividend payout ratio would be consistent with European norms. As I mentioned, the underlying profitability is a bit lower than the headline profitability. To reach a level of EUR 461 million at this stage, a 42% increase brings us effectively to that position.
I think it shows the confidence the board have and the regulator has that that level of payout, that high level of payout is certainly consistent with the performance of the business. That's that box ticked. That obviously allows the issue to move to the next piece, which we know people are interested in, is when can surplus capital come back. As we've indicated, normalizing NPEs, which is the back end of 2019, allows that conversation to move forward. It's a 2020 type conversation. The first two boxes had to be ticked before you can get to the next box. The business is generating 210 basis points of capital on an ongoing basis. We always said that in the long term, 150, 200 basis points of capital generation seem to be the right sort of number. We're continuing to see that coming through.
Donal will go through that in more detail, as will I a bit later on. It gives us a fully loaded CET1 of 17.5% at the end of 2018, after allowing for the dividend. NPE normalisation was a key feature for the business and is a key feature for the business. There is a EUR 4.1 billion reduction in the period down to EUR 6.1 billion of NPEs, 9.6%. Crucially, getting below that 10% number for the first time. I think what you can see is the momentum in the business has continued to go at the same sort of level as we had before. A lot of questions that we have had in the past are, well, when will the momentum stop? I think what we are saying is the machine is built, it is working on a case-by-case basis through all of these restructurings.
Yes, there was a loan sale, but the majority of what has happened in the period has been through restructuring activity in the business. That is going to continue to be the case. The confidence level of getting to the N19 target is high. All of that gives you a return on tangible equity, 12.4%. Again, slightly elevated because of some one-off benefits, but it does allow the business to continue to invest strongly in improving the customer franchise, which is the leading customer franchise in the country. That is macro set of messages for the organization. It is obviously all built on the strength of the Irish economy. Again, similar set of slides and tiles to what you have seen before. On the top left-hand side, you see the overall level of GDP growth in the economy. Again, we reference back to where we were around IPO time.
The net point of this slide is to say growth expectations continue to be elevated relative to the original expectations. The economy, on average, is doing better than people anticipated, that trend has continued at this point in time. It is obvious we will talk about Brexit. It is obviously a threat out there at this stage. When you look at it in overall terms, the economy is performing strongly. On the top right-hand side, you see the employment numbers. That is obviously very important. You can see the unemployment rate collapsing, 5.7%, as we stated at the end of 2018, and 2.3 million people in employment. Getting back to record levels of employment and unemployment coming down very strongly. That provides a very strong consumer base for what is happening across our activity and our customers' activity. Bottom left, obviously, a crucial issue around housing.
Again, a much talked about topic. On the positive side, you can see that the rate of delivery of new housing has increased significantly, which is good news. We also all know that the rate of delivery, about 18,500 units last year, is probably only about half what it needs to be to get to the normalized level, depending on when you are a 30,000 or a 35,000 type annualized demand person. There is also some pent-up demand. The good news, delivery has increased significantly. It was only 15% of demand three years ago. It is now 50% and rising quite quickly. It is going to take a long time to get to that clearing level. The mortgage market should continue to be buoyant for that period of time, the housing stock does need to continue to move forward. On the right-hand bottom, you see that sentiment.
It continues to be positive, not quite as positive as it was a while ago, across all sectors in Ireland and even in Europe, still maintaining an expansionary inflationary trend in terms of overall sentiment for economic activity. What does that mean for AIB? Drawdowns during the period rose from EUR 10.5 to EUR 12.1. Overall level, you can see increased activity across pretty much all sectors during the period. I'll call out one notable exception in a minute. You can see across all aspects of the business, and I'm going to tell you in a while about a change of the business operating model. All of these numbers on the basis of the historic model for the business around the retail, commercial, corporate banks. You can see growth coming in across the board.
On the middle set of tiles, you can see that actually mortgage lending grew by EUR 400 million. Personal lending, although when you round it in billions looks flat, was actually up 5% year-on-year. Some continued growth in personal lending. Corporate and SME, good growth, EUR 4.6-EUR 5.3. Within that, the SME segment has effectively been static and slightly declined during the period. That is the point we would make about Brexit, that we do think there's been a holding off on investment by SMEs as they wait for clarity or to understand where Brexit is going to go. That's certainly a trend we have noticed across the board. More broadly in larger corporates and institutional, they're investing. That's the one piece that didn't move forward in the period was where SMEs were.
Top right-hand side, standard set of market share showing the market leading positions we hold across all key segments. On the bottom right, and Donal will go through it in more detail, I'll just highlight it. A EUR 3.7 billion increase in the performing loan book during the period. Very strong growth in the performing loan book as we had indicated. The strategy on a page slide. I'm trying to get you to the point where you get bored with some of these after a while, which is the strategy is the strategy is the strategy. We cannot to get too excited about short-term issues because we think we have the ability to see through these. Ultimately, the purpose of the organization is to back our customers to achieve their dreams and ambitions. We talked about that a year ago in terms of its launch.
It's now rolled out across the whole organization as embedding. It's based on the values in which we're trying to change and organize the culture. It's putting our customer first, working better together, keeping things nice and simple, empowering staff, and building trust and appreciation both internally and externally. That effectively plays into our four strategic pillars, which is how we invest and how we organize the resources of the organization to deliver for our customers. We do put the customer first, and we just put one stat there. Personal relationship NPS was +35 at the end of Q4 2018. The similar stat for Q4 2017 was +21, so an increase of 14 in the period. Simple and efficient, 64% of key products are sold online or mobile. That was 61%. Some of these are slow builds. They take time. You invest. Customer behavior changes.
64% is where we're at at this stage. It's a slow build. Risk and capital, as we've talked about, generating 210 basis points of capital. Talent and culture, the key metric there is iConnect, which is the Gallup survey, an international survey of how staff are feeling, how they're engaged. We're at the 72nd percentile at this stage. When we started this in 2013, we were on the fifth percentile. You can see, heading towards upper quartile performance at this stage. 17 engaged staff members for every one disengaged. When we started this, it was three disengaged for every one engaged staff member. Quite a transformation in terms of how people in the organization feel about working with AIB, and ultimately that manifests in terms of customer service and behavior, because if you're happier where you are, you tend to deliver better service.
It all does link together. The bottom two boxes are effectively talking about the targets, just reconfirming the targets that we've set out, I'm not going to go through those on the left. On the right-hand side is the base in which we organize to deliver for the customer. I'm going to update that in a minute. Effectively, this is a strategy on a page, no change. If you look at how that delivers then for customers, we've only selected some items here. We've always said that if you deliver for the customer, it'll deliver for the business, so we've given some more NPS, Net Promoter Scores, on the left-hand side across personal homes and SME. You can see, I mentioned the personal up, +35, homes at +50, and SME at +56.
When we started this journey, most of these were at zero or negative, and our target was to get to 50. We've reached those targets. We now need to push on. You can see customer advocacy for what we're doing is strong, and it's manifesting in terms of growth in the loan book, generating the capital, and that new lending staff that I mentioned, 98% at strong or satisfactory grades. Satisfactory is a word that the credit community like, which most people think isn't a great word, but it actually means the PD, the probability of default, on that entire 98% is less than 110 basis points. It's a very strong credit grade for what business we're writing at this stage. In the middle wheel, you can see how customer interactions change over time. Mobile interactions since 2013, almost a tenfold increase.
These are daily interactions, 1.2 million daily interactions. This is a day in the life from 2018. The other one I'd point out is ATMs. You can see cash is gradually eroding itself from the system, 432,000 ATM withdrawals down to 298,000. The only other comment I'd make is on when you look at branch transactions, while it was 77,000 in 2013, and it looks like it's risen in 2018. Actually, 55% of those transactions in 2018 are fully automated, no staff-assisted transactions. It's down in the 40,000. An actual drop in terms of staff-assisted transactions, but people are visiting the branch to use the technology. On the right-hand side, we just called out a couple of user stats to show you how it grows. We got excited, well, I did anyway, between 2014 and 2015 when we went past a million active users online.
We're now at 1.38, a 10% increase in a year. Continued strong take-up of mobile and online by customers. You can see that that shows in terms of customer transactions completed online, up at 44.5 million. Quite strong growth over the 36 million the previous year. Three particular ways we've invested, and one of these is crucial because we'll possibly get a question on mortgages later on. One of the ways we've looked at this has been how do we make the experience better and better for AIB customers? Our core strategy has been increase the conversion rate of AIB customers. That is our core target market, AIB customers, be better for them in everything we do. The focus on the mortgage journey has been to take away all the pain points that customers feel. It's taken a long time.
We talked about it, I think, at the half-year. It's now implemented and rolled out very recently across the whole network. Really, the customer benefits are a fast decision. Effectively, seven out of 10 customers should get a decision in less than 90 minutes, and the mean should be closer to an hour in terms of a decision. You are credit approved within that time period. That's a huge benefit for customers relative to process that exists at the moment. You have a dedicated app, which allows you to upload documents effectively on your own time. You can view your application status at any point in time, and you have constant access to a help center, which is our mortgage center of excellence in terms of dedication to deal with any queries you have.
We're trying to address all those pain points that customers had about feeling isolated and feeling they needed somebody else to solve the process. We should take the pain, make it easier for them. You can see the customer experience, plus 61 on NPS. In fact, 50% of applications we've literally just launched that are now going through this express journey. We want to get up to 70% in its current mode, then push on and get more and more people in through the system as you deal with the exceptions. The MyMortgage adoption rate, the app, is 90% of all new applications. A lot of work. It sounds very dull, but actually has a huge impact on the business going forward. Opening a bank account has always been a problem. New-to-bank accounts were one of the key issues.
How do you deal with verification? A lot of work has been put into that, and now we effectively have a system and a process in place to deal with that. 20% of new-to-bank customers have been opened through that. Literally just launched, we would expect that to build. The key point about all of these is it improves customer experience, but improves the efficiency for the bank. It's a double benefit in terms of how these land. The final one is obviously one that we're all familiar with around payments. Obviously, we've been investing to make sure that we can bring on new payments. Whether it's Apple Pay, Google Pay, Fitbit Pay, we have those.
You can see that ultimately, mobile payments are up 39% in 2018, and over 30 million payments per month, which is a 25% increase, and 50% of those are contactless. About 10% of those are happening through these new technologies. It's a slower build than people might think, but it's there and will happen. Our investment spend, we've always said we invested quite heavily for the period up to 2018, and you can see the box on the left-hand side averages over EUR 250 million. What we're saying is the categories of spend will continue to be the same, but the level of spend will drop. We expect that about EUR 225 million is the envelope of investment in the business going forward. Strategic spend, things I've talked about, MyMortgage app will be there. On process side, I'll just call out Voice ID.
We've just implemented Voice ID for customer and for staff. Effectively all internal process has Voice ID in terms of authentication, again, much more simple and efficient way. If people lose their password, they can now Voice ID themselves back in. For customers, we've trialed for 30,000 customers just to get them used to the idea of using Voice ID to get all the way through call centers right through to authentication simply by using voice. Again, efficient, effective way from a customer point of view and from the bank's point of view. Resilience, we consider this to be payment systems, so some people might call this strategic. The brand new payment system, Dovetail, Calypso, are in the resilience box, well through the implementation phase at this stage, fully completing in 2019. Sustainment is an important one. It's kind of an easy one to miss.
As all of this activity takes place, it changes the way the customer interacts and what you need to deal with. There's been a 1,300% increase in the amount of data we process since 2011. It's doubled in the last two years. You need to invest heavily to handle all this data, to have the infrastructure to support it. You're not improving anything. You're just sustaining your ability to deal with increased volumes. Then regulatory spend is an ongoing piece. What we're saying is expect it all to continue. We expect all these categories to continue to move ahead, but overall, we expect the level to reduce slightly. This is what it looks like when you put a graphic together in terms of new spend, old spend, new systems, old systems, and what we're working on.
I'm not going to go through detail other than to say the way we think about it is customer engagement. That's all new, very important. That's fresh in keeping with what the customer expects, so you invest heavily in that. The integration area the customer doesn't care about. That's our technology that allows us to extract information and to deal with our core technology. We've spent heavily on that as well because you need to access data real time to be able to move forward. The insight layer is what allows you to understand the customer better and to give information back real time to the customer. The record layer is where the hard work takes place. Very boring. Debits and credits, a lot of it. The core system is very resilient, very robust. Some of it's quite old.
We will eventually replace that over time, but it's not really the core issue. The customer sees nothing. We don't see any business benefit from changing it because you've dealt with all of the pieces that matter from a customer and an operational point in the other layers. Foundational technology, very important. Mainframes need to be new, need to be up to speed. The data issues that I've spoken about on sustainment, you need to be able to handle the volume. That's how we think about investment. That's how we think about modernization. The other piece that's changed is obviously how we evolve the business model. The chart on the left you've seen before in terms of embedding the customer focus across the organization, making sure we have that culture.
The piece I'm going to talk to is on the right-hand side, which is the change that took place from the 1st of January 2019 to get closer to the customer. We spent a lot of time thinking about how we could do that and effectively organize the business along 3 vertical pillars at this stage. Customer-based focus around anyone with a business interest, anyone with a home interest, an aspiration to have a home, and anyone who's a consumer who effectively uses digital banking, online banking, branch banking. How do I think about it as a consumer? Most people use at least the consumer channel. If you're a home or a business customer, you probably use both. We've organized ourselves to deliver value from a customer point of view and from a bank point of view across these segments.
We've brought together in this business and customer services group all of the distribution, all of the operations, all of the technology functions, which effectively are melding together to become an operational platform to support customer delivery. That's how technology is moving and enabling us to move forward. It's a cost opportunity and an efficiency opportunity. The verticals are the value opportunity, and all of that is supported by risk, finance, HR, and a corporate and strategic affairs unit, which allows us to think more broadly about the stakeholder set, sustainability, and how we're interacting with the overall ecosystem of politics, of regulation, of customers, and make sure we have that perspective at a group level. That change has happened. It'll be reported on that basis in 2019.
Took a lot of time to get there, and it's supported by what we're doing on property, supported by what we're doing on technology to enable agile working. If I was here next year, I'd be really boring you to death with how wonderful it is. We'll allow that to happen by somebody else next time. The final piece I want to talk about is what happened on capital. As you can see on the left-hand side, Non-Performing Exposures have reduced very materially in 2 years. Again, every time we do this, somebody asks the question, well, are you near the end? Of course, we're near the end. The rate at which reduced is by EUR 8 billion over the last 2 years. The machine is still there. It's working through case-by-case restructuring. We never had a single strategy for doing it.
It's a multiple sort of strategies approach to dealing with it. Very high level of confidence we're going to get to the normalization piece. I'll let Donal talk about that further. That has supported the increase in the dividend. As the balance sheet has de-risked, the dividend has moved forward, and obviously, the capital generation is very important. Starting point for capital generation is really the first tile of January 2018, which is 17%. The IFRS 9 adjustment happened, like for like, you went from 17.5% down to 17%. Generate 210 basis points of capital in the business. We had adjustments, I'll let Donal talk about these, some of those are driven out of AFS portfolio, it's a 70 basis point reduction in the period from that. On a like for like beginning and end, you go from 17% to 18.4% capital.
90 basis points is paid away in proposed dividend, which gives you EUR 17.5 billion at the end, 17.5%, EUR 13.4 billion of capital. For this stage, we've put in the return at the bottom, EUR 10.8 billion paid at this stage, plus 71% of the bank as our remaining holding. I'm going to hand over to Donal, then we'll be back for the Q&A. Thank you.
Cheers. Thank you very much, Bernard. I'm obviously very happy to be here to deliver the financial highlights for AIB for 2018. Also tinged with an element of sadness, given that this will unfortunately be the first and only time I'll get to do some results with Bernard. Genuinely, thank you very much for all of the support, guidance, et cetera, over the last number of years. We will miss you, we know where you will be working in the future. Okay, to the financial highlights. Profit before tax, EUR 1.25 billion, which we would consider a strong financial performance. Our net interest margin is 2.47% for 2018. We did make an accounting adjustment around our net interest margin, this is in the area of how we treat suspense interest on cured loans.
I'll talk a little bit about that later on the income statement, that has had the effect of the change in the net interest margin year-on-year. Costs were EUR 1.4 billion, which was stable year-on-year. New lending, EUR 12.1 billion, which is up 15%. That's a very strong result, that's been across all asset classes, I'll be able to go through that in a little bit more detail. Non-P erforming Exposures of EUR 6.1 billion. Bernard would have mentioned this as well. That's now 9.6% of gross loans, a year-on-year reduction of 41%, we're very happy with that outturn. Very strong work from our FSG group. CET1 ratio 17.5% at the end of the year. Again, strong underlying profitability supporting that dividend payout. We also, in 2018, began our MREL issuance program.
Post the holdco setup, we would have issued 3 transactions in 2018. We feel like we're well on the way to meeting all of our MREL targets by 2021. On the income statement, I'm not going to go through this line by line. Net interest income, stable year-on-year and strong. Other income, we'll come to the breakout of that later, but we're very comfortable with the fees and commissions and the continuity there. Operating expenses, like I mentioned, more or less in line year-on-year. Credit impairments and write-backs, EUR 204 million. Naturally would have guided to you previously that we feel our normalized cost of risk is around 30 basis points. Just given the amount of restructuring NPE reductions that's taking place, and obviously the strong macro environment in Ireland, it has led to another year of credit write-backs. We had exceptional items of EUR 167 million in 2018.
I'll break those out later. Naturally, they're a little bit lower year-on-year. End result is a PBT of EUR 1.25 billion, which we're very happy with. Then just to see here the net interest margin, okay. In the way in which we'll report going forward, you should look at the net interest margin at 2.47%. We've restated the impact of the interest income between 2017 and 2018 so you have a like-for-like analysis. We would always have claimed that our underlying net interest margin was, in fact, this number, and then showed the net interest margin differential. Okay, talking about net interest margin. A few moving points on this. At the start of the year, net interest margin was 2.5%. Over the year, we had a positive benefit in cost of funds of 7 basis points.
That's really in the deposit area. That incorporates any associated cost of holdco issuance, which was approximately 1 basis point. Overall, that would have been a gain. Excess liquidity would have been a drag of 6 basis points for the bank in 2018. Really what that was from the start of the year, the strong macro environment was leading to increased current account SME type of balances. On top of this, we began an MREL issuance program, issuing EUR 1 billion in H1. Then on top of that, again, we obviously had a portfolio sale. Lots of liquidity events, shall we say. It took us a bit of time to consume all of this liquidity and actually make things a little bit more efficient. There was a drag of 6 basis points.
Loan yields for the year, down 3 basis points. Not too much. Wouldn't read too much into that overall. Large mix there. Investment securities, 2 basis points. That's really just the headwinds that we have on our balance sheet from in-the-money Irish government bonds that are just rolling off the balance sheet. Overall for the year, 250 to 247. I'd probably draw you to the numbers here at the bottom. Obviously, at the half year, we could see that the liquidity position was building up. As we move from 250 down to 243, we had to take some actions. Number of items that we would have done, we would have repaid our TLTRO, EUR 1.9 billion. There's no point borrowing money from the central bank if you're giving it back to them at -40.
We would have put in place a deposit pricing strategy with non-bank financial institutions and corporates to either deflect or pass on the cost of funds that we receive on these excess funds, which are negative at 40 basis points. The NIM trajectory is strong, 243, ending the year at 248, which we are pretty happy about. Other income, like I said, fees and commissions, the stable part of the other income. Very stable year-on-year. Other business income, that would include a NAMA subordinated debt coupon, and it also includes where we hold customer-related derivatives or CVAs, which can move over time. Year-on-year, we are still comfortable enough with that EUR 44 million. In the other items, there are a few items there that we see from year-to-year. Gains on securities, reduction year-on-year. That's just really an outcome of portfolio management activities.
Gains on equity investments is a new addition to the other income line, really a new IFRS 9 type of addition. Within that, there are a number of different equities, things like Visa shares revaluation, or other equity positions that we may hold from debt equity swaps, et cetera. Overall, positive outcome of EUR 27 million. A realization of cash flows and restructured loans. That's simply when we agree a restructure with a client, he agrees to dispose of a particular asset, and when the value comes through, it's just in excess of where we had the loan marked. Again, positive valuation of EUR 84 million. Again, kind of tied to the strong environment, strong underlying collateral values. Costs year-on-year, stable. A number of items here. There are a couple of headwinds, which would be wage inflation, which is coming through at 3%.
Secondly, would be the increased cost of depreciation. Obviously, given the investment that we've made in the last number of years, there is going to be a payback impact of that. They're the headwinds. I would say on the different levers that we can pull going forward to manage the cost base, obviously, given the investment in technology, there are efficiencies just to be gained from the business, which we'll continue to do. We also have a number of projects that we've had to stand up for various reasons, whether it's tracker mortgage review or whether it's even the non-performing unit. These are business units that are set up for a point in time, and that we don't think need to endure at the same kind of level going forward. They're the different levers that we think we can pull going forward. Exceptional items here, EUR 167 million.
There are some gains here, such as a gain in a portfolio sale, there are some obvious one-off items that you can see here. I've pulled out a few. Property strategy, which is obviously related to moving out of here, amongst other things. IFRS 9 and other regulatory costs. 2018 was a really big lift from the IFRS 9 perspective, just getting all the models in place and getting the reporting in place. We would expect to see that normalize. We have some redress and restructuring type of costs, which will be associated with things such as the tracker mortgage review. Balance sheet. I certainly am going to go through this line by line, but there are two things that I would like to call out.
What I am trying to do here is give you the walk from the old IAS balance sheet world into the new IFRS 9 world, which I do by just from December, January, and then December. We can see the changes there, and we are talking the same thing. Performing loans up on the year, EUR 3.7 billion, which is an increase of 5 or 6%, which we would see as very positive. We would have talked last year about having reached the inflection point. Obviously now we think we are certainly getting through that. Interestingly, on a net basis as well, even taking account of all the Non-Performing Exposures and provisions, we have balance sheet growth as well, which I think is probably a signal of the normalization of the balance sheet, which we would hope to see continue going forward.
On the liability side, you can see in customer accounts that impact that I would have mentioned to you, an additional EUR 3 billion of liabilities. No real one-offs here, just slow, steady increase linked to the macro environment, we would imagine that that is going to continue going forward. I point out here that there is no ECB funding. Obviously, we repaid our TLTRO of EUR 1.9 billion. At the moment on our balance sheet, there is zero ECB funding. I think that is relevant because in 2010, I think at the height, we were borrowing EUR 37 billion or EUR 38 billion from the ECB. If you were looking for a metric of how things have changed in such a short period of time in the Irish banking system, that is not a bad one. Gross performing loans.
Bernard would have talked on the macro side about the Brexit effect that we are currently seeing, particularly in Republic of Ireland. We do know that SMEs are holding off on investment decisions. We can see that that market is slightly down. We also know that there is a positive Brexit effect. There are more companies moving to Ireland. There is more FDI. Larger corporates are doing larger acquisitions to improve supply chains and get Brexit ready. If you look through the different asset classes, residential mortgages, as we know, slightly up on the year. The market for mortgages overall probably disappointed a little bit at 8.7%. That is clearly going to be a market play. Property and construction, strong year-on-year.
Corporate and SME, large increase year-on-year, and that would really be in the corporate area as opposed to the SME area that I said was relatively benign. The other personal area, which is positive. You can see here on the bottom the makeup of the overall balance sheet versus the actual asset classes for new lending in 2018, and you can see the different moving parts there. Overall, a very positive sign with the increased trajectory in assets. Next, NPEs. Bernard would have talked about the results year-on-year. I just really want to walk you through how we got to where we got to. Always like to remind people, back in December 2013 that the NPEs were EUR 31 billion. We need to put everything in perspective when we are talking about NPE ambition.
At the end of December 2017, the NPEs were EUR 10.2 billion. Again, I am just trying to create the link here between the IAS and the IFRS 9 world, because on the 1st of January, obviously we implemented IFRS 9, and we were an early adopter of the definition of default. In fact, on the 1st of January, our NPEs moved from EUR 10.2 billion to EUR 9.6 billion, with a 33% coverage ratio. As we have said before, the structure that we have in our performing unit is we use three different levers to reduce our NPEs. It is all really focused around trying to engage with customers and get solutions. If we engage with a customer, we get a solution, we make a deal, you will see that coming through the cash redemptions.
If we engage with a customer, we make a transaction, we make a deal, then they will either remain in probation or be agreed for a collateral disposal. You can see here on the left and the right, the probationary period statistics and also the collateral disposal statistics. That box in the middle is quite a fluid number. It has got things coming in and things going out, but going down. Then last is portfolio sales. Obviously, as you know, we would have transacted a large portfolio sale in 2018. That brings us to the endpoint of EUR 6.1 billion. You can see the probationary period assets, collateral disposal assets will have reduced, and the coverage level overall would have reduced to 27% as well. I highlight here on the right-hand side, the December 2019 target of circa 5%. That is a hard target for us.
We are very focused on achieving that. I think with the three pillars that we have in place, we have a strong momentum, and we are very confident that we will be able to reach that endpoint. Sorry, I should say that we do not consider that an endpoint. That is a milestone as opposed to a destination. Sorry, I stole that line from Colin as we were working through the slides yesterday. I just want to unpack the NPEs here a little bit, okay? Because the coverage ratio on NPEs is typically something that deserves a little bit of time. Again, you can see the summary of January 2018 to December 2018, the type of asset class that is in the NPE buckets, and also what the coverage levels are. I think the non-residential mortgage asset class, as you can see, are contracting fairly rapidly.
Really, the questions we typically get are around provisioning, around deeper residential mortgages. What I have tried to do on the right-hand side here is really break out what we have in that book. Okay? You have EUR 800 million of buy-to-let, EUR 2.5 billion of primary dwelling homes, with a provision level of EUR 700 million. I have put here on the right-hand side collateral valuation, EUR 3.1 billion, really just to give you a sense of where things are valued in terms of assets. That is a capped real estate value. Okay? Individually, loan by loan, valued against their specific assets. That gives us a lot of comfort, okay, that we are well provided for. The 20% coverage ratio on arrears PDH book would look low vis-à-vis European peers.
If I just bring you here to the slides on the right, you can see that 36% of the PDH book is in fact not past due at all. That's restructured and performing, sitting in a probationary period. Okay? That asset does not require a significant coverage whatsoever. If you look on the left-hand side, the greater than 180 days DPD, these are longer term arrears. This is the area where you can imagine the heavier coverage levels up around your 50% reside. That's really the picture that I'm trying to show you there. Just over the period of time, the last three years, we put here the loan-to-value on the buy-to-let portfolio and the PDH portfolio, just to give you a sense of how valuations are moving on this portfolio.
No doubt we'll have some questions on that later, but I hope that's clear. Funding structure, an area dear to my heart, naturally. The good news for 2018 was really that we got upgraded to investment grade with all of the rating agencies at our holdco level. Really what that does is it just makes it a little bit easier for us to reach our MREL targets with a lot more confidence. We would have issued three transactions in 2018, two EUR, one USD, all well received. We look forward to more transaction in the future as we go to reaching our target of 28.04%. In terms of liquidity metrics, you can see all very strong. Loan-to-deposit ratio, 90%. Strong Liquidity Coverage Ratio, strong net stable funding ratio. Really that strong liquidity position that I talked about.
Punchline there is that we're certainly very well positioned to finance future growth in our core markets. Risk-Weighted Assets. The outcome for the year for 2018, I would say, was positive. Our RWA density reduced from 58% down to 56%. Within that, there were a number of moving parts. Balance sheet growth and asset mix would have increased RWAs by EUR 1.4 billion, particularly with that heavier corporate lending mix that I would have talked about. Obviously, grade migration and NPE reduction was a counterbalance, netting one another off for the year. Other factors impacting Risk-Weighted Assets for 2019. Off the bat, IFRS 16 is going to have a half a billion EUR negative impact, which would be around circa 15 basis points. Next is the TRIM review.
We've undergone review with the central bank on our AIB mortgage model and also our AIB corporate model. Unfortunately, I don't have anything to report in this area as we have not got any guidance back from the JST on how that's going to land. If I look at domestic peers, if I look at European peers, I would certainly get the sense that it's going to move upwards, but that is just a sense. We have not got any guidance whatsoever from the regulator in this area. If we look below on capital requirements, SREP, I think the main point here is the P2R, 3.15%, so no change year-on-year. P2G, obviously not disclosed. You can naturally assume that there was a marked reduction in P2G as well, just to account for the phasing of the other buffers.
Lastly, I just mention EBA stress tests just to prove that sometimes, from a regulatory perspective, we get positive outcomes. You'll remember our stress test in 2016, where we were certainly at the lower end of the European scale. Post the 2018 stress tests, we were comfortably mid-table, really showing strong progress overall and the progress that the bank had made. Capital ratios, Bernard would have run you through this earlier. 17.5%, IFRS 9 definition of default impact 50 basis points. Strong profitability, 2.1%. Investment securities and other, I can sense Stephen Lang is going to ask me lots of questions about that. Then the proposed dividend payout, obviously 0.9, leaving us at 17.5%. Just a brief summary of the investment securities. That is simply available-for-sale bonds where the benefit is accruing into interest income, and the benefit is taken out of CET1.
I'm happy to take a question on that later, now I shall finish off on the medium-term financial targets. All I'm really trying to do here is paint the picture of what we try to do, which is deliver on our targets, and effectively just do what we said we'd do. Net interest margin 240, outturn at 247, that's obviously strong and in line. Cost-to-income ratio, less than 50%, 53% outturn. Would have discussed the headwinds, I also would have discussed the levers that we have to pull. Fully loaded CET1 ratio 13%, clearly well in excess of that. Return on tangible equity 10%, and the outturn is 12.4%, which is clearly very strong. At the bottom, again, this is the dividend projection that Bernard would have alluded to. We certainly feel at the end of 2018 we're very close to that normalized dividend payout level.
44% payout ratio is very strong. Once that normalized level is reached, obviously we move on to a conversation around return of excess capital. Like we would also have said, that really is a conversation for 2020. I shall now hand you over to my great colleague and soon-to-be boss, Mr. Colin Hunt.
Morning. Thank you very much indeed, Donal. Firstly, I'd like to open by paying tribute to our outgoing CEO. Bernard has been an exceptional leader of our business. That much is obvious from the set of numbers that you will have seen this morning. He has also been an extraordinary guide and friend and mentor to so many of us in the business, I want to take the opportunity now to personally wish him and his family, every success and happiness in the future. The approach to doing business, which has generated the results that we present today, is the approach to doing business that we will continue to adopt over the course of the next number of years.
At the very, very heart of our culture is a customer first ethos, which is now fully embedded in our business through our pillars, also more recently through our purpose. We will continue to augment and protect that ethos over the course of the next number of years by reducing the complexity within our organization, by compressing its hierarchy, and by being increasingly accountable. Our operating model is already evolving. It's now allowing us to align our structure to our strategy, and that's a process that will continue. Thanks to that evolution, we have found ourselves in a position where we can vacate Bank Centre after 40 years here. We're moving to a new corporate headquarters, and we'll also have three new centers of excellence, a center located around our capital city.
As a consequence of that new property strategy, we have to become more agile and modern in how we go about our daily work. That is very much accommodated by the returns on the major investment we've made in technology over the course of the past number of years. That technological investment has not only allowed us to become Ireland's leading digital financial services institution, but it has also allowed us to undertake the property evolution that is ongoing before your very, very eyes. This is all enveloped within a commitment to supporting the ongoing sustainability of our business model, while simultaneously increasing our contribution to the sustainability of the communities in which we live and operate. Prior to the IPO, we set out four ambitious medium-term financial targets. We would maintain a strong and stable NIM in excess of 240 basis points.
We would deliver a robust and efficient operating model characterized by a cost-to-income ratio of less than 50%. We would deliver a strong capital base with a CET1 fully loaded basis of 13%, and we would also deliver target returns to our shareholders on tangible equity in excess of 10%. Those targets have served the bank extraordinarily well and served our shareholders well. They will remain our medium-term financial targets. To conclude, all of the team at AIB is committed to delivering on a consistent and continuous basis for our shareholders. We will do that through an unremitting and relentless focus on our customers. We have an extraordinary market-leading franchise in this country. We will seek to optimize that franchise by always bringing products and services to the marketplace, which allow us to deliver on our purpose.
We do that with the single aim of generating long-term value for our customers and for our shareholders. That said, there is a near-term cloud on the horizon. It's now 28 days away, potentially, and that is obviously the issue of Brexit. We, under Donal's leadership, have been preparing for this eventuality since 2017, and we've been working to ensure that regardless of the events on March the 29th, that we will be able to continue our business operations and that we will continue to support our customers with advice and with capital. For the year ahead, for 2019, our number 1 focus is on normalizing our non-performing exposures. That will allow us to withstand any future economic shocks, and it will also allow us to deliver very attractive capital returns. Thank you.
Okay. That's the end of the formal presentation piece. We're going to open to Q&A now at this stage. What we'll probably do is take a couple of questions in the room, then we'll give people on the phone an opportunity to come back with any questions they have, and then we'll finish off in the room again with another question or possibly two. That's the general flow we'll try and keep, and we'll see that Eamonn had his hand up quickest. Off you go.
Thanks very much. Eamonn Hughes from Goodbody Stockbrokers. Thanks for that, Bernard. Maybe if I can just move on to the AFS point since it came up a number of times, maybe how we should think about that in terms of prospectively as well, if that's an issue in relation to capital accretion into maybe 2019 and even 2020. Maybe just bring us through the kind of the components to it, possibly. Secondly, just in relation to NIM progress, I think it's slide 14, you gave kind of a good helpful walk in terms of the pluses and minuses. Maybe is there any sense you can give us in relation to how that walks through in 2019?
Finally, I know you mentioned kind of no update on TRIM for now, but should we kind of expect it falls in 2019 in terms of the impact, or could it drag out a little bit further?
Okay. Well, you've solidly targeted Donal in those series of questions, and obviously there's a few gaps in your model that you need filled. We'll be able to answer some of those, but probably not all of those. Want to start with AFS?
You can see the 70 basis points reduction in investment securities and other. Within that, there is a change of 54 basis points, which we'll call an AFS effect. Okay. Of that amount, two-thirds would be related to in the money bonds simply accruing into interest income. Into interest income and out of CET1. A third approximately would be due to the weaker credit market in Q4 of 2018, which would have had an adjustment in that area. If you're looking for what could happen in 2019, I would say that could be 30 basis points on CET1. That's the AFS question. The net interest margin one is a little bit more straightforward. Obviously, the exit NIM is 248. In terms of cost of funds, we think that that will actually be flat. There's obviously MREL headwinds.
Some of this is dependent on where the market is, because we have to issue. We have a target to reach. We think that we can end the year flattish on cost of funds because there's some benefits to be made on the deposit side. Excess liquidity we don't think is going to be a feature in 2019. Loan yields, I would put in as stable. There's obviously stuff rolling off, which is on lower yields and stuff coming on at higher yields.
We probably guide flat on that. The AFS headwinds will remain at two to three basis points per annum. That was very clear, I think. Lastly, the question was around TRIM. We would have completed the mortgage TRIM review in 2017, the corporate model review towards the end of 2018. We just do not have any visibility or line of sight on how that is going to land. I really would be very reluctant to give you any color there, because I just don't know.
I think the only thing we'd say from a sort of continuity of what we've said on this topic before, because we have nothing new, is you know where we start in terms of RWA density, which is higher. We have no indication that where we start from is usually problematic from that point of view. We never anticipate good news coming out of these things, we just can't give you anything further on it. We obviously start from a different position than many others started from. Yeah. Stephen.
A couple questions from me, Stephen Lang from DB.
You might just wait. There's a mic coming in of your right shoulder there. Sorry.
Very strong growth in the wholesale institutional business, again during H2, significant growth year-over-year. Wondering if you could break that out a bit more and also provide some sort of clarity on the evolution of that. Is it sustainable to expect similar year-over-year growth? Presumably it's at a lower base, but anything in that regard would be helpful. Then secondly, Donal, you reiterate there that you've previously talked about a more normal cost of risk of 30 basis points, but we've just come out of a period where strong write-backs again in H2 as well as H1 and on the prior year. How do you see that evolving into 2019? Thanks.
I'll just make one comment before we pass over to Donal on the cost of risk. Then maybe something else on the wholesale side. The cost of risk is always this question of the long-term position. The long term happens in the long term. We're obviously coming from a positive position at this stage. We're in 2019 already. We've made the observation around the strong credit quality that's been written at the moment. I think the statement that we make is true, because it's our best view as to where that'll be in the long term. In the short end, it's probably likely to be more positive than that longer-term position. I think that's been a statement we've made before. I think we're still in that position at this point in time.
Obviously, there are external factors that may have an impact on that, and we talk about Brexit as one of those examples. There are events that may change that, but on a regular flow basis, I think that's where we're at. I don't know if you want to add anything else to that, Donal.
I think on the write-backs specifically, the change we made on the treatment of the suspense interest. Effectively, what it meant is EUR 44 million in 2018 came out of interest income and would have gone into the provision write-back line. Okay? Supporting that. Obviously, we have a fairly aggressive NPE target, there is going to be more restructures happening throughout 2019 as we reach the target for the end of the year. It's more than likely that we'll have positive write-backs again in 2018, but I don't think it's going to look like what it was, sorry, for 2019, but it's not going to look like what it did for 2018.
We've invested very heavily on the corporate institutional side over the course of the past number of years, we are seeing the rewards of that investment coming through in terms of balance sheet growth. We had a very good performance from the Syndicated & International Finance team, that was matched by a very strong performance in particular from our real estate finance team. Right the way across the businesses, all parts, all units within what was WIB are performing very, very well, I'm happy to say that the early indications are that we've had a good start to 2019.
Okay. We'll take one more in the room, then we'll go to the phone at that stage.
Cheers. Pierce Byrne, Cantor Fitzgerald. Two questions. First one just in relation to NPE levels and the excess capital distribution. Around the time of the IPO, 5% and the European average NPE ratio would have been used interchangeably. We're probably seeing that European average drop below probably 4% at current levels. Do you think would there be any issue with the ECB in terms of approval of a capital expenditure if AIB was at 5% and let's say the European average is sub 4%? My second question is just in relation to the funding side. Your own deposit ratio fell to 90%, we're seeing it as a continued headwind for NIM.
Can you give us any kind of guidance in terms of the actions you took during the year to limit or to pass through the effects of negative rates and we saw it in kind of the context that we saw the loan book or the deposit book grow by EUR 2 billion. Finally, just have you any thoughts on what effect a possible TLTRO could have if the ECB announced one later this year? Thank you.
Well, I'll leave the last two questions, the LDR and the TLTRO for Donal to pick up. I'll just make a general comment on the NPE point, and if Donal has anything else to add, he can. It was an inevitability that when everyone targets improving to the average, the average improves. It's just math. I think what the ECB did was they stated that position and then effectively agreed targets with each individual institution. That's just the formula and the process that has happened. Our target for the end of 2019 is based on what we've agreed through an ECB process around that normalization piece. There's nothing new or surprising has come from that. I think Donal has already made the point that we knew it wasn't the end of the game, but it was the end of the normalization period at that stage.
If you also think of the charts that Donal has shown, while the NPEs at that point in time may be at sort of a 5% type number, you've actually got different components in those, some of which have already been cured and are in probationary periods. Actually your core unresolved NPEs is actually significantly lower, which isn't the case for other jurisdictions because they don't go through the workout solution. You need to look at the characteristics of NPEs at the end of 2019, not just the number. I think it goes back to the point that we said that the conversation is based on getting to that point, and then effectively you need to look at these as a series of discussions that take place around excess capital.
I don't think anything has changed from that point of view at all. I think the only thing we could say is that relative to a stretching target, we're ahead of where that was and are comfortable that we have the momentum within the business to get there. The other two for Donal.
Yeah. I would have highlighted in one of the liquidity slides that we have no ECB funding. The reliance necessity requirement for AIB on ECB funding TLTRO clearly doesn't exist. I think you were probably going what that is, if TLTRO is withdrawn or et cetera, is it going to have an impact on markets?
More so just if the ECB try to use it as a stimulation tool, will you be able to draw on that, and will it have any effect? Will we see -
Okay. Yeah. Okay. Listen, we've repaid the TLTRO. We were getting money at minus 40, and we repaid it. The loan-to-deposit ratio at 90% is at that level because the core liabilities continue to grow. Now the assets are growing. If you want to manage an excess liquidity problem, grow your balance sheet. I think that's what we're showing we are doing on a gross and a net basis. That's why I'm saying going forward, I don't think that that's going to be a feature.
Okay, we're going to go to the phones now.
Your first question today is from Alastair Ryan from Bank of America. Your line is open.
Thank you. A number if I could, please. You still have very high Pillar 2 requirement now that the bank's completely different to when you got that. Is there any hope of that coming down? This theme that the competition have drawn out in the last few days about needing to get to 100% coverage of problem assets by the mid-2020s, regardless of whether those are actually performing restructured and what have you. You've rightly brought attention to the EUR 3.1 billion in mortgage collateral you've got. Does that end up not mattering? No, I'll leave it at two, please. Thank you.
Okay. Well, again, we'll do a double header on this one. I think your comments about Pillar 2 and where we are from a regulatory capital point of view, let's take a broad statement around them. We have always said that until it's demonstrated in the balance sheet, the regulator doesn't give particular credit to it, which is obviously a key difference to the equity market. I think we've always been consistent about that point when people were saying, "Well, you're very close to whatever that target might be. When do you see something happening?" We've always said that it has to be in the balance sheet before it happens, and I think that statement is proving to be true in respect to things like TREP, which is actually off 2017 balance sheet.
We know that there's a lag between delivery or expected delivery and when it actually gets manifest or realized in a regulatory conversation. However, we obviously have a high dividend payout ratio this year, which is evidence of the progress that's been made in terms of the conversation with the regulator and their comfort level about distributing capital. Donal may have something more specific he wants to make you aware of, but I think our view is we do have a high one. We obviously have a business that's performing well, is changing quickly, and we expect that over time that'll manifest. For now, we can't give any commitment as to when that'll happen. There's nothing different or new that we know. It's just that issue of takes time for it to actually come through, and we think that's still the case.
Yeah, nothing really to add. Probably fair to say that the main items on the regulator's radar would be things like non-performing exposures, which is certainly not a surprise to anyone. That's why we have such a comprehensive plan in place, and we have really clear targets in place. Then vis-à-vis any JST guidance, I think there's clearly lots of talks of letters being sent throughout Europe, which is wonderful. The way we look at is that the fact the regulator is concerned about NPE's coverage ratios, et cetera, does not come as a surprise to us. We have been in the middle of a fairly comprehensive program of NPE reduction with the regulator for quite a period of time. The guidance for us is really this 5% number by the end of 2019.
Obviously, there's going to be an evolution of the strategy, and obviously that update is going to incorporate answers to questions and solutions to any future regulatory guidance that we would have received. In summary, we'll get back to you on an update.
Yeah. I think, again, the metrics that Donal produced, if you go back a little bit further than that, the residential portfolio was over EUR 8 billion. It's now at EUR 3.3 billion. If you look at the portfolio that's past due, it's only half of that, and that's well provided. If you do all the maths on that, you end up talking about a relatively small portfolio in an unprovided category. The letters that people are referring to is only consistent with the guidance that came out earlier. It shouldn't be a surprise because they came out with the guidance and said this is what they're going to do, and then they started writing letters.
We think the strategy we have for trying to get through those portfolios is working. We think before any of those may or may not become effective, we'll have made significant further progress anyway. None of that is totally surprising from our point of view. We think that perhaps we have a benefit that we were at this earlier than others. We certainly have the competence, the capability, the machinery built to get through it. The track record, even based on what happened between 2017 and 2018, of a very significant reduction in that portfolio during that time period from restructuring. We think all of those elements will play out. We know that it's a piece of news that people are discussing. We think that everything we have in place actually allows us to be more comfortable on this topic.
Thank you. I will steal a third if I can, just while you're still with us. Going back to the time of the IPO, there's some portfolios in your book, mainly I think the EBS portfolios, which still got very high risk weights on standardized models. You've had model approvals in for model requesting for approval for a long time. You always guided they would take quite a long while to come through. When we're thinking about TRIM, do you get any benefit from things going the other way as well? It's not just an increase on your IRB models, you might get benefit from standardized moving to IRB. Really that's too optimistic for our point of view. We've got to take the hit on one and maybe the benefit from another could come down the road, you can't see it.
Well, I link it back to the IPO comment. Then if Donal wants to give an update. I think at the time the IPO and in subsequent discussions, what we said was point to point, we see overall that we'll be able to hold reasonably static RWAs, which will get through a growth period and also get through a normalization of the level of standardized we have versus IRB. Obviously that means increasing the amount of the portfolio from low 30s up towards 80, effectively from standardized IRB. Then there is some growth. We see benefits coming from all of that. That's still the case. However, we've said we anticipate that bad news will be front-ended and good news may be back-ended, therefore the transition will be lumpy.
The fact that we haven't heard back on a portfolio, you can overread that. It may be that there isn't too much bad news in it's not important to get there quickly. We really can't give you anything of substance other than think that generic comment holds. We have improvements coming from the move towards IRB. There are also negative moves that'll take place in that period. Then obviously you have the NPE benefits that are coming in over time.
Yeah, nothing to add.
Thanks very much. Thank you.
Thanks, Alastair. We have another call, yeah?
Thank you. Your next question is from the line of Andrew Coombs from Citi. Your line is open.
Good morning. If I could have one follow-up to Alastair's question and one fresh question. The follow-up would just be looking at the disclosure you do provide on slide 40. You're looking at EUR 3.3 billion of residential mortgage NPEs who are still running at 10% run rate. If we're thinking about this point of 50% coverage by end 2020 and then moving to 100% over time, when we're thinking about the potential impact on your 2020 capital position, should we just be thinking about a rundown of that NPE ratio towards 5% and then an increase in the coverage ratio from 20 to 50? Is that just far too simplistic a way to think about it? That'd be my first question.
Second question, looking at the NIM sensitivity you provide on slide 46, can you just provide us with an idea of how that EUR 211 million splits out between the asset and liability side of the balance sheet? Linked to that, any thoughts you have on the impact of any potential deposit tiering by the ECB? Thank you.
Okay.
Okay. I'll take the first question here on the PDHs. Again, we do not want to get into giving any clear guidance on what may or may not be on the balance sheet in 2020 and beyond in the NPE space. We're focused on delivering the circa 5% NPE target at the end of 2019. That will incorporate all asset classes. Really what I was trying to do by giving you this breakout on the right-hand side is to show you what is in fact a deeper rear portfolio, what is not, and what the value of that collateral is, and ultimately what the provision levels are around that. If you put those together, I think you'll be able to draw your own conclusions. We're really focused just on the NPE target of circa 5% by the end of 2019.
Any changes to that, any evolutions of that, we'll come back to you on. The next question was around NIM sensitivity.
Yes, please.
Was that just vis-à-vis loan yields and cost of funds, assets and liabilities, I think is what I heard?
It's exactly that. Also in terms of anything you have parked with the ECB at the moment in terms of your excess deposit base.
Cost of funds overall would continue to see similar trajectory in core liabilities, really in current accounts at zero levels. There will be MREL headwinds, like we said, additional MREL to issue this year, two, perhaps three deals. That's going to be a cost, but we think that there's also some benefits that we can still continue to reduce in the deposit area. That's why I'm really saying flat on cost of funds. In loan yields overall, what we're seeing is a stable environment for all of our main asset areas, mortgages, corporate, SME, et cetera. That's what probably gives me the comfort to guide on a flat basis. In terms of excess amounts with the central bank, this is a euro problem for excess liquidity, okay? Because putting cash in deposit of negative 40 basis points.
We do have some cash on deposit in the U.K., which we would not consider significant or a drag because we get the BOE rate. I think at the height in 2018, we perhaps were EUR 4 billion or EUR 5 billion in deposit with the ECB, and it's much lower now in around EUR 1 billion, EUR 1.5 billion. That's a number that we're able to manage quite comfortably. I think those are the real levers, and I don't see the excess liquidity becoming a problem again in 2019.
Okay. We're beginning to run out of time now because we have a hard finish. We'll take one more on the phone, and then I'll close with one more in the room. On the phone.
Thank you. Your next question is from the line of David Wong from Credit Suisse. Your line is open.
Yeah. Good morning. Thanks for taking my question. I just have one quickly. I think going back to the guidance of, I think, flattish loan yields and flattish funding costs. I guess on the loan side, I guess given the fact that you expect new book lending to grow, and I think one of your competitors also sounded reasonably optimistic about the potential for mortgage repricing, or at least over time. I'm interested why. Could you perhaps explain to us a bit more the reasoning behind your sort of flat loan yield guidance for 2019, please? Many thanks.
Well, quickly, we haven't made any comment about pricing in respect of mortgages, nor do we ever. Our overall view is that the net book, in terms of how it'll expand, will be broadly flat based on our expectations for how that'll develop. Hence our comment on yields is based on the overall portfolio and our position, and the fact that we don't make any statements about pricing until they happen. That's our best information and guidance at this stage. Okay. Obviously those who missed the call, you have your IR contacts, so please get onto us with any detail and for follow-ups. Thank you. Last one in the room over there.
Good morning. Owen Callan from Investec. Two quick questions. One on cost to start with. It looks like you're still investing quite a lot of money in the intangibles, presumably the software side of things, which I know you said EUR 225 million would be the predictable amount of investment going forward. Is that maybe slightly higher than where you maybe would have expected that to be a couple of years ago? Have technology investment requirements increased? Obviously that will start to feed through into an eventual depreciation and amortization charge increase. Given that there's underlying wage cost inflation in the economy that we know about, which you've mentioned previously, does that make it quite tricky to manage the cost line at an overall level, going forward?
Then, on SMEs quickly, obviously with Brexit, we know the SME sector is difficult and uncertain at the moment in terms of credit demand, and it looks like there's been a small reduction in the Irish SME sector. You don't break it out necessarily by SME, but looks like there's also been an increase in the U.K. lending to SMEs or to non-property business certainly, and that might be a little bit unexpected versus where consensus or where the models were. Would that be something we'd expect to see a continued increase in U.K. lending, until at least the Irish stuff kicks in? For 2019, given Brexit, what is your outlook for the SME sector?
Okay. The quick answer in respect to the investment side, we have been calling up the investment piece since probably 2013, 2014. That's why we're guiding on the new numbers so that we're clear about where it is. I think you're going to have to invest across the business, and that's the reality of where it is. What we do see on the other side, what Donal had alluded to earlier was, there are other benefits that come out of that, the efficiency operating model. That's why we tried to give that perspective that while there is customer benefit coming out, there's also operational benefit coming out. If anyone ever tells you managing a cost line is easy, it probably isn't. The items you're referring to exist. Inflation 3% last year, pretty much where we thought it'd be.
We know those things are there. You have to manage to them. They are difficult, they are tricky, we know it's a key focus point. I don't think much has changed, to be honest, in respect of that. I don't know if Donal wants to make any comment on the SMEs, but I'd say overall, the U.K. isn't an expansionary market for us at this stage. While the new lending was up, the balance sheet was flat. We wouldn't guide people to think about the U.K. as an expansionary market in 2019.
Nothing to add.
Okay. What I might do at this stage is just make a closing comment. This is obviously my last one of these. It's been a tremendous honor to be Chief Executive of the bank. I'd like to thank all of you for most of your interactions. I did have the joy, with many of you, of the first outing in 2011 when we talked about the 2010 results. It was a EUR 12 billion loss at that stage. The good news is we've moved on quite a bit. Definitely sad to be leaving the bank after nine years, but after nine years it is time to move on and to hand over to others who are fresher and have more enthusiasm at this stage for the great things that can happen. I'd like to thank my Chairman in particular, for his unyielding support.
Obviously we've acknowledged that Mark has departed as well. Many would have commented that we were an interesting and good double act. As a result, I'm kind of delighted to be able to leave the podium and the leadership of the bank to Colin and to Donal, I think they're going to be a superb team. They're great colleagues to have had during my time here, I think they will be a great double act as well going forward. Thank you very much, I wish everyone great success. Thank you.