OSB Group Plc (OSB) Earnings Call Transcript & Summary

Good morning, everybody, and thank you for taking the time to dial in to our half 1 2020 results presentation today. When we put our full year 2019 results out, we were in the group have never before witnessed full lockdown with rapidly rising levels of COVID-19 infections as well as associated deaths. None of us really knew then whether or not the world was going to hell in a handcart. But now, whilst we don't have total clarity on the future outlook, there are definite green shoots and, I believe, more room for optimism. During the somewhat strange period, we have focused on our colleagues and our customers, enabling a significant majority of our employees, in both the U.K. and in India, to work from home. Where that's not been possible, we've adapted our office locations to provide a safe, COVID-secure working environment. Our branches have remained open throughout the pandemic to serve our savings customers as they required. And our collections and servicing teams throughout the group have worked tirelessly to help borrowers facing potential financial difficulty through the arranging of mortgage repayment holidays, 26,000 of them as at the end of June. Our sophisticated and conservative approach to credit and risk management has served us well. Our LTV on the entire group portfolio is 66%. We have high interest coverage ratios across both OSB and CCFS buy-to-let books. And our underlying impairment charge of GBP 54.4 million is driven predominantly by the adoption of more severe COVID-related macroeconomic scenarios which, we believe, place us at the conservative end of the spectrum versus industry benchmarking. We have been very focused on capital management and balance sheet optimization. We entered this pandemic with a strong CET1 ratio and have strengthened it further in the first half to 17.4%. Clearly, we are now making use of the Bank of England's TFSME scheme. And we have been accredited to grant loans under the CBILS scheme should we see opportunity or if the scheme is extended. We have successfully integrated our capital markets team. And in March this year, we undertook our largest securitization to date of circa GBP 1 billion, which we have retained to substitute the AAA notes with the Bank of England, the TFSME funding, at significantly lower haircuts than all loans. As of the 30th of June, the group's liquidity coverage ratio was 245%. Immediate outlook terms, we remain focused on supporting our customers, colleagues and communities. Our improving mortgage application volumes are approaching 60% of pre-COVID levels but at higher pricing and tighter criteria. On the application levels we're now seeing, we expect to deliver double-digit full year net loan book growth with NIM broadly flat to the first half. This slide contains our statutory numbers. However, as the comparative is against OSB solo, I'll now share the highlights of the group's underlying results so we can show you more meaningful comparatives to last year. Gross new lending in the first half was GBP 2.1 billion which, I believe, is a strong achievement given the elongated hiatus of COVID lockdown and the inability to even undertake physical mortgage valuations. This has enabled us to deliver 7% net loan book growth, stripping out the impact of structured asset sales that we delivered earlier in 2020. Our NIM remains attractive at 250 basis points. And whilst this is slightly down on the equivalent prior period, due in part to running through the pandemic with higher liquidity for prudence and the fact that it took the retail savings market longer than in normal conditions to pass on the large base rate reductions to back book and front-end pricing, this margin still represents a strong return on secured lending. April will give you some more color on our NIM shortly. Having clearly been able to spend less through lockdown on things like travel and recruitment, for example, we have delivered an exceptional cost-to-income ratio of 26% in H1. And despite increasing our IFRS 9 provisions significantly due to those worsening macro scenarios, we have delivered a return on equity that many banks would kill for in a normalized market of 18%. Our profit before tax is down 14% due to those increased provisions but still a healthy contribution of GBP 156 million. I'll now hand you over to April to give you some additional detail on our financials.

Thank you, Andy, and good morning, everyone. Before I take you through the results for the first half, I just wanted to give you a quick reminder of the key differences between our results on a statutory underlying and pro forma underlying basis. As Andy mentioned, statutory results include results for the first half of 2020 for the combined group, but the comparatives for the first half of 2019 are for OSB only as previously published. Underlying and pro forma underlying metrics reflect results for the combined group as if the combination has closed on the 1st of January 2019. They also exclude exceptional items, integration costs and other acquisition-related items. We believe that these provide a more consistent basis for comparison. So let's start with return on equity. I'm delighted that we still delivered an exceptionally strong high double-digit underlying ROE of 18% for the first half despite recognizing significant loan losses under the IFRS 9 approach. Underlying profit before tax, as Andy mentioned, was GBP 156 million versus GBP 183 million in the prior period. And this was after a GBP 54 million loan loss charge as we adopted more adverse COVID-related forward-looking economic scenarios, which resulted in an underlying loan loss ratio of 60 basis points in the first half. Underlying net interest margin remained attractive at 250 basis points, albeit down on the prior period, and I'll explain that in more detail on the next slide. Before we do so, I wanted to talk about our continued focus on cost discipline and efficiency, which improves our cost-to-income and ManEx ratios even further in the first half to 26% and 68 basis points, respectively. We kept a tight control on cost during lockdown and also saw the benefits of synergies in the first half, mainly from senior role deduplication. Looking forward to the full year, I would expect the cost-to-income ratio to increase marginally versus the first half due to the impacts of the gain on structured asset sales in January, which increased other income in the first half. The next slide shows the evolution of underlying net interest margin over the last 12 months. We start with a NIM of 270 basis points in the first half of 2019. You can see the dilutive impact of the changing asset mix of the OSB loan book as it refinanced onto front book pricing. This has largely run its course by the end of the first half, but the exit rate going into the second half was lower than the first half average. This drag was partially offset in the second half of 2019 by the positive impact of EIR adjustments taken at 2019 year-end. These were nonrecurring, leading to a reduction in NIM in the first half of 2020. You can then see the dilutive impact of the delay in passing on the base rate cuts in full to retail savers and the higher liquidity we prudently ran with -- which impacted the second quarter. This resulted in the first half NIM of 250 basis points. Looking ahead, based on current pricing, I would expect full year NIM to be broadly flat to the first half, with the dilutive impact of the delay in changing saving rates and the higher liquidity carrying on well into the third quarter. The retail savings market has largely passed on the base rate cuts by the end of June to new deposits, but there was a further delay in passing this on in full to our back book of easy access accounts. The current outlook for Q4 NIM is positive, and I expect it to be broadly flat to fiscal year 2019. Turning to the income statement. I'd like to highlight the gain on structured asset sales in the first half of GBP 33 million on an underlying basis. This included transactions in both OSB and CCFS, an early success of integrating our capital markets functions and demonstrating our continued capability to accelerate organic capital generation through the sale of residual certificates. Fair value losses on financial instruments included losses of GBP 14 million in the first half due to fair value movements on mortgage pipeline swaps prior to the mortgages completing due to a fall in the outlook for rates. We economically hedge our pipeline but can't start hedge accounting until the mortgage is complete. However, this unrealized loss will unwind over the life of the swap. We generated underlying earnings per share of 26p per share in the first half, down on the prior period due to the impact of the higher impairment losses. The chart at the bottom of this slide provides bridge between underlying and statutory profit before tax, the main variances being the amortization of the fair value uplift on CCFS' net assets recognized on combination and a lower gain on structured asset sales as a result of this uplift. Now let's turn to our strong and secure balance sheet. Total originations were GBP 2.1 billion in the first half, down on the prior period due to the impacts of the pandemic, which particularly impacted the second quarter. This new lending drove 2% underlying net loan book growth in the first half, or 7% excluding the structured asset sales. As Andy mentioned earlier, the property and mortgage markets have recovered quickly since lockdown restrictions were lifted, helped by a range of government stimulus. And based on current application levels, we expect double-digit net loan book growth for the full year, excluding the impact of the structured asset sales again. Our balance sheet remains primarily funded by retail savings, with diversification provided by securitizations and Bank of England funding schemes. We took steps to prudently increase liquidity at the onset of the pandemic, drawing GBP 645 million under the Bank of England's Indexed Long-Term Repo scheme in March. We also attracted strong retail inflows during the ISA season. The group remains highly liquid during the first half with a very strong liquidity coverage ratio of 245%. We were accepted for the TFSME scheme with a combined initial allowance of GBP 2 billion, with additional allowances expected based on net lending year-to-date. We expect to use TFSME to refinance and extend the duration of the original TFS and also our ILTR drawings with the potential to draw down further amounts to fund growth opportunities. Our capital markets team placed GBP 847 million of RMBS bonds into the market during the first half. And as Andy mentioned, we issued our largest securitization to date under the Canterbury program, a wholly retained deal which provided us with GBP 860 million of AAA-rated bonds which are eligible as collateral for commercial repos and central bank funding facilities at significantly reduced haircuts compared to the underlying mortgages. We exercised strong diligence over loan and customer assessment throughout the first half, tightening criteria as the pandemic took hold. Our loan book is secured at sensible loan-to-value, a weighted average of 66% at the end of June and 68% for new lending during the period. Affordability remains strong for buy-to-let lending with interest coverage ratios on lending in the first half above 200% for both OSB and CCFS. The credit quality of our loan book remains strong, with 3-month plus arrears stable for the first half of 2019 at 1.3% for OSB and 0.5% for CCFS. The next slide provides more detail on impairment provisions, which drove the underlying loan loss ratio of 60 basis points in the first half. The graph shows the evolution of impairment provisions since 2019 year-end, the main increase being the impact of adopting the more adverse COVID-19-related forward-looking economic scenarios, which almost doubled the year-end provision of signals in our Q1 trading update, with a loan loss charge of GBP 42 million. We made enhancements to the staging criteria for payment holidays, in line with regulatory guidance, which moves certain higher risk accounts to stage 2, attracting a lifetime probability of default, leading to a loan loss charge of GBP 5 million. The remaining increase in provisions was due to some modeling enhancements and business-as-usual seasoning of the loan book and the impact of new lending. The business-as-usual loan loss was equivalent to 6 basis points loan loss ratio, consistent with our broadly stable arrears position. You can see the significant increase in our coverage ratios as a result of adopting conservative forward-looking economic scenarios with the overall coverage ratio more than doubling to 51 basis points. So let's take a look at our new forward-looking macroeconomic scenarios on the next slide. The table shows the projected annual movement in GDP, unemployment and house price growth between 2020 and 2022. The HPI assumptions are particularly conservative even in our upside case. Let's take a look at payment holiday behavior on the next slide. At the end of June, we had nearly 26,000 accounts in payment holidays, representing 28% of the loan book by value. The graph shows what has happened to this cohort up to the 14th of August. It was apparent at the time of granting these payment holidays on a self-certified basis that many of the borrowers were requesting them to prudently safeguard cash flow rather than as a necessity. This has been borne out by the small percentage of extensions for maturing holidays of 16% by account number and 18% by value. These extensions represent about 5% of the June loan book by value. It's also reassuring that we continue to see low levels of requests for new payment holidays since June. This behavioral data is comfortably within our first half IFRS 9 modeling assumptions. Turning to capital. You can see the group's very strong capital position at the end of June, with a fully loaded CET1 ratio of 17.4%, total capital ratio of 18.6% and a leverage ratio of 6.8%. The graph explains the improvement in the CET1 ratio over the first half. The cancellation of the 2019 final dividend added 0.6 percentage points. The Capital Requirements Regulation Quick Fix package added a further 0.5 percentage points. This included 0.4 for the capital relief increases in IFRS 9 stage 1 and stage 2, expected credit losses and 0.1 for the extension of SME support factor. The structured asset sales in January added a further 0.8. A combination of the gain on sale and RWA relief, the amortization of the fair value uplift on the CCFS-acquired loan book with a drag of 0.3 and, finally, profitability of the foreseeable dividends, growth in the balance sheet and other items provided a net decrease of 0.2 to arrive at a 30th of June ratio of 17.4%. Turning to MREL. We received our annual resolution letter from the Bank of England in July, setting out their preferred resolution strategy. As expected, we are subject to single point of entry bail-in requirements which, from July 2023, is expected to be 18% of RWAs, risk-weighted assets, rising to an end state of 2x Pillar 1 and Pillar 2a by July 2025. We intend to fulfill our MREL requirements through senior debt issued via a new holding company, which we are in the process of setting up. Our first issuance is likely to be in late 2021, subject to market conditions. I'm pleased to report that we have made good progress on our IRB project since integration, and we are planning to submit module 1 to the PRA next year. We remain of the view that achieving IRB will be beneficial to our capital requirements, especially under the new standardized calibrations and final IRB output floors as outlined in Basel III. I'm just going to finish off with a quick review of our segments. Post combination, we report under 2 segments, OSB and CCFS. All of the segments have been impacted by reduced lending through lockdown and by higher impairment charges due to adopting the adverse COVID-19 forward-looking macroeconomic scenarios, so I will not repeat these drivers each time I move to a new segment. Okay. Taking each segment in turn, we continue to segment OSB further between buy-to-let SME and resi. So I'll start with the buy-to-let SME subsegment, and the loan book was flat in the first 6 months at GBP 9 billion, with new organic origination down 38% at GBP 0.9 billion versus GBP 1.4 billion in the same period in 2019 and due to the structured asset sale in January. Net interest income in the subsegment increased to GBP 230 million, up 8% versus the prior period due to a higher average loan book. The buy-to-let SME subsegment made a contribution to profit of GBP 111 million, broadly flat to the prior period, with the higher impairment charge offset by higher other income, including the gain on structured asset sale. Average loan to values remained sensible for both stock and new origination at 68% and 71%, respectively. Turning to OSB's residential subsegment, the gross residential loan book increased by 6% in the first half to GBP 2 billion driven by the first charge gross loan book growing by 9% to GBP 1.6 billion despite lower organic origination in the segment of GBP 185 million, down 29% on the prior period. Net interest income from the residential segment increased 9% to GBP 34 million on a higher average loan book, and contribution to profit was broadly flat due to the higher impairment losses. Average loan to values remained low at 58%, with an average LTV for new lending at 69%. Turning to the CCFS segment. All numbers are presented on an underlying basis, and I'll talk to the 2 main subsegments. The buy-to-let subsegment gross loan book grew by 5% to GBP 5 billion after structured asset sales with strong new origination of GBP 1.1 billion, down on the GBP 1.5 billion in the prior period. Net interest income was down slightly to GBP 56 million, and contribution to profit reduced by 11% to GBP 50 million after the higher loan losses. The average loan to values remain sensible at 74%, with new lending at 73%. The CCFS residential subsegment gross loan book on the next slide increased by 4% to GBP 2.2 billion, with new originations down 37% to GBP 237.2 million. Net interest income was up 8% to GBP 34 million versus the first half 2019 on a higher average loan book. However, contribution to profit was down 23% to GBP 23 million after the higher impairment losses. The average LTV remains sensible at 70% with new lending at 71%. I'll now pass you back to Andy.

Thank you, April. When we announced the merger of OSB and CCFS, part of our industrial logic was to create a larger, stronger group, more capable than each individual firm of either maximizing opportunity or managing headwinds. I genuinely believe we are stronger together, and that has aided our navigation of this pandemic. Our integration is on track, including the delivery of cost synergy benefits expected in the first full year following the combination. And I'm really pleased with how much integration project planning and work has progressed through this period with teams finding new ways to work together to deliver results. In light of the pandemic, the Board has chosen to review some of our longer-term integration activities, such as our strategies to move where practical to fully centralize through our single supplier operating models and instead of seeking to keep some strategic flexibility in order to maintain the best in operational resilience. We do not, however, expect this to materially impact the delivery of our cost synergies. We've already seen the benefits of sharing best practice and mixing our talent pool across the group. I've already mentioned some of the benefits realized through the integration of our capital markets function, enabling enhanced balance sheet optimization across the group. However, one of the more challenging areas in any merger is to combine and define a single culture for a new group without losing the strength and uniqueness of either. In a funny way, COVID has helped us accelerate that cultural journey. Culture is often centered around the office location, with some feeling closer to certain parts of the action than others. But with so many of our staff working from home, the location tag disappears, and teams have quickly come together under common management to work across the group for real mutual delivery. Our pulse surveys and our regular communication events confirm that people are really, really starting to feel like 1 team. This, coupled with our strong savings and lending franchises and our approach to robust risk and balance sheet management stand us in good stead for the balance of our successful integration. Our award-winning lending franchises continue to be well regarded by broker and borrowers. Our lending is prudent with interest coverage on buy-to-let originations at 203% in OSB and 205% in CCFS. Our borrowers through the OSB Choices program continues to refinance with us above our target retention rate, with 69% choosing to take a new product with us rather than refinance away. And our proposition continues to serve the professional end of the buy-to-let market with professional landlords accounting for 94% of OSB completions in half 1 and limited companies making up 52% of CCFS buy-to-let applications. As you would expect, we sought to derisk in the current uncertain period with LTV offerings and criteria are tighter in our core markets. And whilst we remain open for business in our more cyclical lines, such as commercial bridging and development finance, we have, as I'm sure you would expect, significantly reined in our risk appetite. The group remains predominantly retail funded, and our strategy remains the same as it always has been, to attract new savers for the value for money savings and ISA accounts, to retain those savers through fair treatment and pricing and to deliver excellent customer service along the way. I'm pleased to say that we've done exactly that in half 1, attracting 22,000 new savers to the group, retaining 95% and 86% of maturing savers in OSB and CCFS, respectively, and delivering very high Net Promoter Scores across both banks of 67% and 83%. Of course, balance sheet optimization and wholesale funding is also a key strategic deliverable, and we executed structured asset sales, generating GBP 33 million on an underlying basis and our largest securitization to date during half 1 2020. So how are we feeling about the outlook? Well, the forecasted potential HPI declines are not -- certainly not at the moment forthcoming. In fact, the housing market has been stimulated and is functioning extremely well. Application volumes continue to bounce back and at those higher prices and tighter criteria, and we're seeing very encouraging experiences with the payment holiday exits. Our integration is well on track, and our cultures are coming together really nicely. Our risk and credit management is strong. Our macroeconomic scenario and modeling is prudent. And our risk performance in H1 is stable and marginally improved than the equivalent period last year. And our LTVs are sensible at 66% on stock and 68% on the flow. And we remain well capitalized and highly liquid. As we've said, we do expect to deliver double-digit net loan book growth for the full year 2020, excluding the impact of those structured asset sales. We expect brick NIM to be broadly flat to half 1 and the cost-to-income ratio to be marginally higher versus the first half due to higher other income in H1 from the gain on asset sales. But most importantly, the group is a well-oiled and resilient machine, and we are capitalized to deal with further problems should they occur or, in fact, maximize the opportunities that are likely to come as the economy starts to normalize. Thank you for listening, and we'll now open up for questions.

[Operator Instructions] Our first question comes from Benjamin Toms from RBC.

I've got 2 please. You've noted in your presentation your macroeconomic assumptions on house prices remain very conservative relative to peers. Can you give us some color on your thought process here? Are you seeing something that others aren't seeing? Or is it that you felt directionally it would be wrong to recalibrate these assumptions upwards at this point in time? And can you give us some idea the order of importance of GDP, unemployment and house prices in driving impairments? And secondly, you've disclosed that 28% of your book is on the payment holiday. And for loans rolling off the holidays, only 18% are going on to another holiday. Can you break that 28% between different types of lending? Or are you seeing similar trends across the whole book?

Thanks, Ben. I'll take the assumptions one and then ask April to touch on the repayment holiday piece. I mean, yes, I think it's evident that our macros are relatively conservative versus the market. Obviously, we've looked at all the benchmarking of peers and other organizations. We've looked at the Bank of England scenarios. And I think as a Board, we just concluded, that's what we've used to arrive at a provision number. We don't have enough long-term data under our belt yet to absolutely assess that we should be winding back from that. And it doesn't feel like the right time at this stage to be unwinding it. So we've maintained the prudence and start with it and our auditors have been supportive of that. I don't think it's because we're seeing anything that others aren't seeing. I think we just have a desire to be conservative in the way that we run the organization. The prioritization around GDP, unemployment and HPI, I mean the 2 big things that affect arrears and then potential losses on arrears for a secured lender is the unemployment rate and then, obviously, the value of the security after that. So HPI and unemployment are the ones that we focus a lot on. I guess GDP decline is an indicator that those things could be in the offing, but that's our focus. April, are you happy to touch on the repayment holidays?

Yes. I mean, just to echo your point on the drivers. I mean unemployment is the big driver of probability of default. And then HPI is a big driver of the loss you take on default. Yes, I guess we haven't published breakdown, but we have, I think, told you in the past that there's a slightly higher take-up of payment holidays in our near prime resi book than the prime resi book or buy-to-let books. But it's not dramatic. And I think you would probably expect to see that. And there is clearly a correlation between payment holidays and self-employed as well. But I think those are probably the main things to draw out.

Our next question comes from Ian Gordon from Investec.

I've got 3, if that's okay. Firstly, unlike most of your peers, you continue to take a more hair shirt approach to your accounting presentation for fair value losses on financial instruments. Like you say, these wash out to 0 over time. So could you please, firstly, just remind me the average period over which the 18.6 H1 '20 should reverse out? And then secondly, do you have a cumulative total of negative fair value losses which will be subject to a similar positive reverse? Secondly, kind of following up on Benjamin's question on your economic models. As you say, the assumptions you have, especially in relation to HPI as well as unemployment are considerably more conservative than some we've seen elsewhere. I guess this question is a bit of a statement of the obvious. But given that your underlying impairment charge of 6 basis points is a rounding error, can you confirm that, subject to your economic assumptions being ballpark correct, one would expect the impairment charge in H2 '20 to be considerably lower than the charge we've seen in H1? And then thirdly, I just wonder if you could give me a bit more color about your pricing strategy and your risk appetite in some of the higher margin areas. So first of all, can you just give us some commentary on how are you seeing pricing development in the core buy-to-let market? And then secondly, what would you need to see to step back into areas such as development finance which are clearly very positive to NIM?

Yes. Thanks, Ian. And I'll probably ask April to tackle the first 2, and then I'll talk to you about pricing and appetite in the other lines.

Okay. Well, I guess, I'm trying to be transparent in what's driving that fair value loss. As with a substantial pipeline of 2-year and 5-year mortgages, we are having to hedge that economically, but the accounting rules don't allow you to start hedge accounting until the mortgage is complete. And with the -- when the LIBOR and SONIA curve outlook declines, we tend to take a loss on it. I mean the pipeline swaps for 2 years and 5 years, so sort of -- I would probably imagine somewhere like 3.5 years weighted average life, just top of my head. And there's actually a note in the accounts we can draw your attention to afterwards, which gives you the balance sheet. It's the hedged asset and hedged liability, I think, but we'll help you find that after the call. And then on the impairment charge, yes, I mean I agree with what you said. If we have perfect foresight on the economic outlook and that comes to pass and we have all of our probability of default and our roll rate assumptions in our modeling 100% correct, then what you'd expect to see going forward would be a business-as-usual run rate. One word of caution, of course, is our business-as-usual run rate and the new lending will be attracting the higher probability of defaults because we introduced those more adverse economic scenarios at the end of the first half. So it would be a little bit higher. Having said that, I think it's an interesting point. What do you do? If you have a portfolio of loans and you've assigned 10% probability of default to that cohort because they're similar. When the 10% defaults, you would increase your probability of defaults for those loans. But would you feel comfortable about releasing and lowering the probability of default quickly for those that haven't defaulted? Or would you wait a little bit of time? So I guess there's always the possibility that you may decide to be prudent as you see the defaults and not kind of assume everybody else is fine. But in the absence of that, I agree with your remarks.

Thanks, April. Pricing and risk in high margin areas, I mean, you touched on in core buy-to-let what are we seeing. So I mean, we have pulled LTVs in, so the maximum LTV that's available from us is 75%, whereas it used to be 80% and 85% in some occasions. So we pulled the LTV criteria in. But what we didn't do is reduce the pricing at the same time. I think it's fair to say that the OSB brands are less price-sensitive and precise mortgages. And that's because it tends to be more of the portfolio proposition, potentially the more complicated end of the market. And therefore, there is a bit more price elasticity in that because borrowers are paying for our service to look at their portfolio in the round. But I think it's fair to say that we don't think that there's suddenly going to be a rush of competition because while the wholesale funding markets for the nonbank lenders are functioning, they're probably not functioning as opposed to they were pre COVID-19 and quite as well. And that makes it a little bit tougher for the nonbank funded competition. And obviously, we're not really competing with the cheap high street rebate. So we see the prognosis on buy-to-let pricing to be good. The higher margin areas, commercial, development, finance, et cetera, I mean, we are open for business. We're writing very little commercial real estate secured lending. And I think we need to get a bit more sense around what's going to happen on office space. We've never been a big fan of retail anyway unless it's sort of Tesco Express with a couple of flats above it. But sort of out-of-town light industrial units still continuing to perform well with lots of businesses functioning. So we're keeping a very close eye on commercial real estate and thinking about being selective around the kind of real estate that we want to lend in. We have written some new development finance deals in the last few months, generally with borrowers that we worked with for a number of years and we know, and they've got good cross collateralization against government schemes. They're in areas where typically they're building houses not apartments, and the sales ratios are going well. Again, there are changes coming up on the help-to-buy scheme. We need to see how that factors in. But effectively, we will just watch cautiously. And then when we think the time is right, we'll start to let the lead back out again. But I just don't think there's enough data under our belt yet for us to become -- hone in those -- in those other higher-margin markets.

Our next question comes from Edward Firth from KBW.

Can I just ask you about the NIM and just check if I got the point of the messages right, but also just ask another question about MREL? So I think I heard correctly, the Q4 NIM, you expect to be broadly in line with 2019, is that right? And is that broadly what we should see as the sort of normalized kick off rate, if you like, or run rate as we go into next year? So I guess that's my first question. Or is there anything we should be expecting next year, which might put that off? But I can't really see it particularly. That was the first question. Second one is you've given us some numbers around MREL. Could you just give us a bit more clarity in terms of what are we talking in terms of pounds, millions of instruments and some sort of expectation that you may or may not have in terms of -- I mean obviously, it depends on market conditions, but what sort of incremental cost will be in terms of NII to cover that MREL? And then I guess, related to that, how does that sort of compare with your -- or how would you manage that against your over 17% Core Tier 1? I mean is there sort of an expectation that you'll be issuing MREL and then reducing your equity in terms of your Core Tier 1? Or how do we -- how do you think about how those 2 trade off against each other?

Okay...

Thanks, Ed, and all complex ones, so they're yours April.

Yes, you did hear me right that Q4 will be broadly flat to fiscal year '19. There was obviously the bit of the benefit in fiscal year '19 of those EIR adjustments we disclosed at the time. But other than that, yes, flat. And well, I guess that's assuming, of course, that current pricing, current spread of retail savings to LIBOR on your base rate continues. I think we have a potential for a bit of benefit, and we're seeing that in Q4. We would expect a higher proportion of our funding coming from the Chief of Bank of England. You've got a little bit of a mix of that funding, which is helping. And I guess what headwinds are there? We've got NS&I out there raising an awful lot of money for the government, uncertain as to when they will have filled their boots. But they are -- despite being government-guaranteed, they are offering higher rates than we are for easy access. So at the moment, we're concentrating on 1- and 2-year bonds. We like those because they're longer duration and they match our balance sheet better, but it's just at a slightly higher cost. So there may be some headwinds there from the government raising funds. But other than that, I think it's a broadly positive outlook for NIM. And if we see green shoots, as Andy described, next year that would make us want to turn back into our cyclical businesses with a little bit more anger, then this may be some positive from margin there as well. Turning to MREL. We haven't disclosed that level of detail. You're quite right that you can meet MREL through CET 1, through AT1, through senior debt. And what we're doing now is looking at the time horizon where we would look to optimize our capital stack. Right now, we're focused on keeping strong CET1 given the uncertain outlook. But I think I'm not prepared at this stage, apologies, to give you that longer-term view on capital management because it will depend to some extent on what the outlook looks like as we go into next year and the year after. But you're quite right that if you've got a particularly strong CET1, it does reduce the amount of senior debt that you have to issue. But I guess I'd reiterate what we said in the prospectus and at the time of the combination, which is our GBP 22 million of cost synergies should more than cover a reasonable expectation of the coupon of the end state MREL senior debt issuance.

Okay. But just as a follow-up, if I may, I mean in any sort of normalized world, it looks like you're making returns of only 20% plus. You're starting with a 17.4% Core Tier 1, you can't possibly grow into that. I mean you must be starting to think something about what to do about that, aren't you?

Yes. I mean, look, we've -- in the past, I've kind of walked you through our capital strategy. And clearly, it's to have enough but not too much capital. I think it's probably a little premature to start thinking about returning capital. I would mention though that the CET1 ratio does assume the foreseeable dividend, the 25% dividend for this year. And we have connected to come back to the market at the end of the year with our intentions in that respect. So yes, I mean, we will look at the outlook. We will look at the growth opportunities, and we will certainly look to optimize our capital stack. And that may, at some point, include share buyback or extra dividends. But I think it's a little premature in this particular COVID situation to draw conclusions or give you a sense of scale or timing. But it's definitely something we think about.

Our next question comes from John Cronin from Goodbody.

And so I want to ask on IRB firstly. I note the statements this morning, the fact that you're going to submit your -- or you're on track or expecting to submit your module 1 application in 2021. Not looking to over analyze that, but clearly you have ongoing dialogue with the PRA in that respect that you set for. And thinking about, look, with the wave of work coming down the road from a PRA perspective of the hybrid models, should I -- would it be fair to interpret that as you're likely to be ahead of the queue as it were -- without any promises on timing, of course, but certainly, you will -- I guess would you expect that could potentially be approved within a reasonable, for example, an easy month time frame if all went well in terms of the robustness of the application. That's my first question. Secondly, getting back to the impairment charges, just thinking forward in terms of life post furlough, could you give us a sense of the level of work you've done in terms of the profile of your borrower base? And I guess that question is more geared towards the residential component of the buck, given the difficulties in establishing that with only a great degree of accuracy from an underlying tenant base of buy-to-let. But we've seen a lot of the banks including yourselves committing to high level economic scenario, views or scenarios rather with unemployment numbers. But clearly, this crisis impacts the customers in the service industry more particularly. Can you give us a sense of the comfort you have around the work you've done in relation to provisioning with that in mind? And anything that you -- any of your expectations in respect of the deployment for repossessions on a longer-term view as well and given the particular nature of this crisis. And then finally, just if I could ask you to comment more broadly, maybe this one is for you, Andy, just around the -- your expectations in terms of demand factors on a medium to longer-term view and despite the less demand, and how does that inform your thinking with respect to the kind of weightings that you would like to see attached to the various loan? How you would like to see your loan mix evolve with -- given your views in terms of the structural evolution of the price of that market now? It does seem to me that there are some very positive drivers there.

Thanks, John. And April, you want to take on IRB and impairment?

Yes. Sure, happy to. So I don't know whether we're ahead of the queue necessarily but certainly ahead of some. And it's a very important question because, of course, the PRA have finite resources. And you don't want to be submitting the application along with loads of other people. But they do manage that. So you submit when they say you can, and that's not just about you being ready. It's about them having the resources. But we've taken the time sensibly since we combined to make certain that we take the best of both application projects, both IRB projects, get the teams together. We've got some super smart people. I don't know if any of them are listening in to this call and blushing. But the key thing for me is that we have a really credible, really strong application when we do submit it, and then we would hope to enjoy that 18-month period that they signal. So I think there's the quality of that first application, which is important and having all your ducks in a row and having that very honest and open ongoing dialogue with them. And I think that's the secret to a successful application process. And then going on to the impairment side of things. Yes, I'm glad you mentioned the end of furlough because I probably should have mentioned that in my other answers because that clearly is the big unknown. You get to the end of October, what kind of customer behavior will we see, not just perhaps on the payment holiday cohort but on the wider book. But as you've seen, we've got pretty tough assumptions for unemployment over the next couple of years, which definitely provides for that in our IFRS 9 modeling. But what do we know about our ball is a lot. I mean we've mentioned to you in the past that we do a lot of forward-looking analysis with the credit bureaus on all of our customers on a monthly basis. We know about their personal indebtedness, we know about their credit score. We know if they're a buy-to-let landlord, if they've got a room or a property up for rent and at what price. We know if they've got up for sale and at what price. And we also now have a lot of information from current account behavior, who is it who's seeing a drop in income over the last few months? Are they the same people who've asked for a payment holiday? And then when we look at our IFRS 9 modeling, the enhancements we made to our staging, and we put 26% of the payment holidays into stage 2, we're looking whether or not they had been in arrears over the past 12 months even if they were up-to-date at the end of June. Who had lower credit scores? Who had higher personal indebtedness? So we very much look at the nature of the borrowers to determine a cohort that we felt were high risk. And then what we're doing is we're measuring that against people who've asked for an extension, is the correlation there? And those of the cohorts as well, of course, we'll be looking at when it comes to potential issues at the end of furlough. So yes, we keep very, very close to our borrowers. And we're looking for all sorts of correlations. Clearly, as I mentioned, the obvious one is the self-employed, which are largely, of course, in our resi portfolios. So yes, we watch this very carefully. But I would say today -- and it is very early days, of course, we are well in size from what we're seeing in the behavior, the assumptions we've modeled, but there is that checkpoint at the end of October. And we probably won't have much insight by the time we do our Q3 trading update. But whatever we have seen, we will, of course, share at that time.

Yes, absolutely. So I mean, in terms of the demand and sort of mix evolution of the book, I mean, I think we're quite happy with the mix that we currently operate, who we like. We like the buy-to-let market, and we particularly like the specialist buy-to-let market. And we also like the residential market, but it's not quite as remunerous on a standardized basis as buy-to-let. That's why we probably write more in the buy-to-let cohort than others. I mean demand, the rental market is just as hot as the sales market. I mean right move, so July was the hottest month on record for a housing market for a decade or activity for them for a decade. Zoopla have just reported, I think, this morning that that's continued -- in fact, I think you've read that message, I think continued into August. And actually, we talked to a lot of the biggest agency chains, Connor is our valuation provider, and they're telling us that the rental market is extremely hot. I mean I always think in these situations, if you think about the lockdown period, there'll be a number of things that have happened as a function of that. But there are a few more people that have broken, I would assume. I think there will be a few more people that are separating or divorcing. All of that will put pressure on a housing market that actually wasn't building enough pre lockdown, had a period off during lockdown, and it's going to take some years to catch back up to being able to cope with population demand. So whether you rent your bricks-and-mortar roof above your head or whether you buy it, the long-term demand prognosis has to be strong.

Our next question comes from Grace Dargan from Barclays.

If I can just ask you a couple. So the first one is just a quick clarification, picking up on the capital point earlier. I don't know, would you be able to add any color on the CET1 ratio that you would aim to operate at over the medium term? And then the second one is on asset sales. And just whether you have any visibility on asset sales and potential gains in the near future and what your current appetite is and whether that's changed at all in the current environment?

Okay. April, do you want to talk about...

Yes. They were difficult questions, so I was assuming I would pick those up. No, only joking. No, I mean I think we haven't given guidance. And hopefully, my -- on the CET1 ratio we was aiming for, I mean, we have in the past given a minimum other than my commitment to make sure that we do optimize that as we move forward based on the economic outlook at the time. And clearly, we have a relatively small Pillar 2a. We disclosed it at 1.67. That excludes the small integration add-on. But -- so you can work out with the capital conservation buffer. And when the kind of cyclical buffer comes back, you can kind of work out what our requirement is. And I think I'll direct you to our Pillar 3 disclosures as well for more detail on that. We will always run with a margin above that for growth opportunities and for prudency. But clearly, in a normalized environment with a normalized outlook, we would not be running at 17.4%. We would certainly be letting that reduce. And on the asset sales, I mean, we view these as capital generation trades to a large part. And with a very high capital, we're really not needing to generate capital and pull forward profits in the short term. But I've been amazed at how resilient the securitization market has actually been, very quickly recovering post lockdown and even during lockdown. So there certainly are opportunities, I think, for us to raise funds through RMBS, not at the same rates as the Bank of England funding, of course. So TFSME offers a cheaper way right now of fulfilling our wholesale funding appetite. But it's very encouraging to see how resilient the RMBS markets is in the sort of appetite for residual.

[Operator Instructions] We have a question from Gary Greenwood from Shore Capital.

I just wanted to come back on the dividend. You said that you'll review that at the full year. So when I look at your financials and, obviously, a lot of upside your capital strength, but also you're very profitable in development despite the pressures that you've been under. I mean it doesn't seem to me that there's any financial reason why you shouldn't be paying a dividend. So I just wonder if you could talk about the sort of the considerations that you have to take on board in terms of being seen to be doing the right thing from maybe a political perspective or regulatory perspective. And then also whether there are any restrictions at all from having taken any sort of government support, whether it's, I don't know, using the TFSME scheme or whether it's having sort of used furlough at all, I don't know whether that's the case. But if you could just talk generally around that, that would be helpful.

So Gary, first of all, let's pick up on government schemes. So no, we haven't furloughed anybody or applied for any of the other sort of GBP 1,000 a head incentives or anything like that. The only thing that clearly we'll make use of is TFSME, but I don't view that as a government support scheme. That's a Bank of England funding tool designed to aid lending into the grid economy. And so, no, we haven't partaken in any of those. And as a balance sheet business that was spinning off income regardless of whether we write a new case every day, it wouldn't have been the right thing to look to furlough. And actually, our staff have been pretty actively engaged and busy throughout the period anyway. So we didn't really have an option to do that. But we wouldn't have it and the Board discussed that in depth in the early stages. I mean, I think on the dividend, there's lots of considerations. We weren't told last time around by the regulator not to pay unlike the large institutions. But clearly, in a scenario where the uncertainty was so great, and it was huge, you do everything that is within your power as a Board to make sure that you've got as much capital under your belt as you can have for a real unknown unknown at the time. And as we get to a point where we've got a bit more clarity and our forecasting is more robust, and we've got more customer behavioral data underway, clearly, the Board has a duty of care to absolutely consider if we're making good profit and we're generating good capital, then clearly we want to make sure that shareholders were rewarded for that. So the Board will have a robust discussion around that at the end of this year and then decisions will be made. Operator, do we have...

Yes, we have a couple more, sorry. We have a question from Robert Sage.

Okay. I was just going to say, we are at 10:30, which was the time, but I think we do have time to do a couple more if you've got them on the system so please carry on, sorry.

Okay. Robert Sage from Peel Hunt.

Just a brief one, actually. And it's in connection with some of the government initiatives, which I think are talking about asking landlords not to evict any of their tenants. And I was just wondering how you sort of see this issue, whether you see it as being very much simply a short-term phenomenon, whether you built it in any way into your provisioning models, whether it affects your pricing decisions, or do you think it's something we can safely ignore?

Well, I mean, clearly, having to go down the process of evicting a tenant is complicated. There are some costs attached to it. So we can't simply ignore it. It is part of our thinking about the collections process or receiver of rent process, et cetera, which you have to take a buy-to-let proxy into possession. But that moratorium comes to an end at the end of September, along with the moratorium on lenders to repossess residential property. And I'm assuming that any landlord who has had a tenant who hasn't paid and doesn't get into an arrangement to sort of get back up state with a landlord, those landlords in October will commence the process for eviction and they have a legal right to do that. Clearly, the government were able -- while the furlough scheme is in place, to be able to say you can't do it because most people have access to income, and even self-employed had access to some level of income back from the government to tide them over. But there comes a point, both from a lender, residential perspective, and a landlord-tenant perspective, where that moratorium comes to an end and things have to move on. So it does come to an end. I don't know whether you could ignore it. You have to think about it and factor it in. But yes, we -- our understanding is it ends and then landlords with proper terms can go through the normal core process.

Okay. We have another question from Aman Rakkar from Barclays.

I'm actually surprised it's not been addressed on the call, but just a question regarding costs. So obviously, the control in the half was really strong. To me, it looks like your costs are going to be down. I know they're probably going to be up in the second half, half-on-half. But they do look like to me there's a decent chance they can be down year-on-year this year. So I mean it would be great if you are able to kind of confirm that. And then secondly, when we're thinking about 2021 costs and thereafter, how should we be thinking about that? Is it a case that we should expect it to grow below your kind of loan book growth because of the synergies that you'll be driving from the integration. So maybe you could do double-digit loan growth next year, mid-single-digit cost growth? I mean anything you can kind of help us to kind of understand that cost evolution in 2021 and beyond would be really helpful.

Do you want me to take that one, Andy?

Well, I was -- yes, I do, but I was going to dive in and just say cost-to-income ratio is clearly a metric that everybody spends a lot of time focusing on. But of course, there are 2 very big halves to that, one of them being the income side of the equation. So your cost-to-income ratio gets skewed when you generate additional income, such as a gain on a structured asset sale. Or actually if your income goes down, that has an impact in terms of cost-to-income ratio. So we also, as a business, focus very hard on our management expenses ratio. And as we have been growing and as we get synergies and benefits of scale through the combination, our natural expectation is that our management expenses ratio will fall to an exceptionally efficient level. And in fact, April had a slide in the deck earlier with our ManEx ratio, and there's been a really good reduction in that in the first half, probably a little bit more than we could achieve in the second half given that we are sort of back-hiring for the vacancies and back on sort of the growth train in a bit more anger than we were during the lockdown period. But we are very focused on cost management through the management expenses ratio, and the benefits of scale that come as a function of the merger and the associated synergies are all helpful in that direction. April, what have I missed?

Well, I think just maybe a little bit more color on what happened in the first half. I mean you alluded to this earlier, Andy. There were some things that just reduced cost without us having to really do anything, if we're honest, because in lockdown, people weren't traveling, they weren't entertaining, and there was less marketing spend because we weren't lending. So you have that. I mean none of these things I'm going to tell you are huge, but they all add up. We reduced our expectations for discretionary bonuses in the first half given the results. And despite the results being incredibly strong, they're not as strong as we had originally budgeted for. And that's what drives remuneration, the original budget. Some of our long-term incentive plans which we have assumptions as to what percentage of them will vest, that expectation was reduced as well. So there was a release of that. So there were some things which, I guess, were one-offs that impacted the first half, and you would expect to not see that benefit again in the second half. And we will start to hire again. So in the first half, we weren't hiring despite the balance sheet growth that you saw, which was a bit of a benefit as well. Having said that, looking more medium term, we do expect higher synergies as we signaled in the prospectus, and you'll have those numbers, and the run rate sort of synergies increasing as we go through the next couple of years. So there's some positives to come there. And as Andy mentioned, we do get the benefits of economies of scale. Not all of our costs are variable. So as the balance sheet grows, the variable costs grow, but the fixed cost base doesn't grow at the same extent. The way we typically manage our costs is we do actually look at cost-to-income ratio because we know it's an important metric to the market. And therefore, we do try and manage our costs to a particular cost-to-income ratio. So when we look at prioritizing projects and spend, we have one eye to that. But nevertheless, during lockdown, we still continue to invest where we needed to invest. We continued to spend a lot of money on our IRB program using external consultants and experts. That kind of important investment in the business did not stop. It was the discretionary spend that was unnecessary in the circumstances that we didn't spend. So I hope that gives you a little bit more color. But I would expect second half to be bigger just because we will have some more people, and you won't have that kind of reversal of those remuneration items again. And we will gradually start to travel a bit more, we already are to some extent, and to spend more on marketing to support the application levels that we're currently seeing. I hope that helps give you a bit more color.

No, that's perfect. Just a quick final clarification on that. So when you're looking to manage cost to income, are you typically targeting something below 30? I mean would you do 28 sustainably, which is probably what you're kind of pointing to in H2? Or basically, is 26 far too low? I mean is it some number in between that and 30? Is that kind of...

I wouldn't disagree with anything you've said. I think you're thinking about it the right way.

Okay. Operator, any final questions outstanding? I think we've got time for one more if there is, otherwise...

No. That concludes the questions.

Okay. Well, to those that asked and are listening, thank you very much for attending this morning. IR obviously is available to fill in additional blanks and so forth. But hopefully, you can enjoy the rest of your day. Thank you very much.