Good morning to those joining in the U.K., good evening to those joining us from Australia, and welcome to Virgin Money UK PLC 2020 interim financial results presentation. Before David and Ian take you through the presentation, I just wanted to briefly introduce myself to those who don't know me, following my appointment as the group's new chairman this week. I have to confess to a certain sense of deja vu here, as I became CEO of Alliance & Leicester in July 2007, at the start of the last major crisis, having been CFO there for six years. I remember all too well being questioned by some of you who are on this call today back then.
I have worked in financial services for over 35 years, and as I mentioned, have the scars and learnings of managing through the global financial crisis some 13 years ago or so. I joined the board of CYBG, now Virgin Money UK, as deputy chairman in 2015 and have been here for much of that time. In addition to Virgin Money, I currently chair Ashmore PLC, the emerging markets asset manager, and I am a NED of PayPal Europe. At Virgin Money, since the crisis began less than two months ago, the leadership team, led by David, have done an astonishing job of reinventing the way we work as we prioritize the health of our colleagues, the health of our customers, and the health of the bank.
The board have been heavily involved in the bank's response to this pandemic, with deep dives into our portfolios, regular appraisals of customer behavior, and the support we are providing. As well as good debates about how we best support our colleagues at this time. The board and I believe we are well-positioned for the challenges ahead. I hope to meet many of you in person in my new capacity as chairman when we emerge from lockdown. For now, I will hand over to David and Ian to talk in detail about our response to the pandemic and our first half performance. David? David, are you on mute?
Sorry. Yes, apologies. David, thank you. I'm delighted to have you confirmed in the role of Chairman, and I think your experience and the continuity you provide will be invaluable, and I look forward to continuing our existing excellent working relationship. Welcome to the presentation, everybody, and thank you for taking the time to join. I hope you're all keeping safe and well in these difficult times. Given the COVID-19 implications, we're going to take a slightly different approach to our presentation today. I will share a brief overview of the business, and Ian will provide a detailed review of our balance sheet strength and portfolios and the first half results. I will finish with some closing thoughts on the implications of COVID-19 for 2020 and beyond.
That should take us about 40, 45 minutes. We'll then pause and go to Q&A, and hopefully be closing it around 10:00 A.M. Before I begin, I also just wanted to update you all on the search for our new CFO. I'm pleased to confirm that we have undertaken a thorough process and have recommended a candidate to the regulators this week, and we expect to receive feedback from the regulator in due course. Turning to slide five. The COVID-19 outbreak, it's clearly transformed the whole world, our industry, and our business. We are confident that the strategic approach we outlined at our Capital Markets Day in 2019 remains appropriate. Let me talk you through our current COVID-19 priorities before returning to the next steps in our strategy.
Our immediate priorities, no surprise, are focused on the health, safety, and economic well-being of our customers, colleagues, and the communities in which we operate. Equally, it's critical that we protect the bank's capital. Given that priority of focusing on our capital, I am confident that we start from a position of strength. We have a conservative balance sheet mix, a resilient capital base, and a prudent funding and liquidity position. This means that we're well-placed to be able to provide the necessary support to all of our stakeholders whilst managing the bank through this crisis. My leadership team and I are working closely with both the U.K. government and UK Finance to help design and deliver funding or forbearance programs to support the various individuals and businesses in the U.K. that need our support.
We're now offering a wide range of support packages to all of our customers who need those. For our mortgage customers, we have granted around 60,000 mortgage payment holidays, which is about 15% of our customers. We have implemented a digital request service to speed up access to this support for our customers. In our personal division, we are proactively supporting our lending customers with, to date, 32,000 credit card payment holidays, which is around 2% of customers. 8,000 personal loan payment holidays of around 6% of our customers. Helpfully, 90% of these payment holidays are delivered online. For any deposit customers in difficulty, we also offer access to their funds penalty-free where needed. Although to date, we have not seen many of those requests.
In our business division, our focus is on supporting our existing customers and extending facilities to ensure wherever possible, viable businesses are supported through a period of what is really cash flow challenges. This can take the form of lending facilities, obviously capital repayment holidays and overdrafts, as well as the different government initiatives that are in place. Our experienced relationship managers have moved quite fast to engage with customers. We've had about 10,000 conversations with our business customers in the past few weeks. To date, we have supported businesses with around 4,500 lending support facilities, including GBP 135 million of lending approved via the government's Coronavirus Business Interruption Loan Scheme. We are also committed to supporting the Bounce Back and CLBILS schemes too. From an operational resilience perspective, we've been able to keep 95% of our branches and all of our call centers open.
Across the business, these call centers are operating at near pre-COVID service levels. I'm enormously proud of how our colleagues have risen to the current challenges, and I'd like to thank them all for continuing to offer excellent support to our customers. I have personally held CEO calls or videos with almost 4,000 colleagues over the past four weeks to thank them, to provide reassurance, and to listen to their concerns. Today, more than 70% of our colleagues are working remotely. Our prior investment in our colleague IT platform has meant that we've been able to implement this new model within 2 weeks. We are helping our colleagues to adapt to working from home. It's interesting that many are telling me that they are more efficient and can offer better service to customers in this new model, which shows great potential for the future.
We also recognize that at this time, our communities need extra support, and our Virgin Money Foundation has made over GBP 850,000 of funding available to local charities responding to the COVID-19 pandemic. Our Virgin Money Giving team has also delivered record donations to charities despite being in lockdown, and this includes having helped raise GBP 13 million in donations to the NHS and over GBP 20 million last month to all charities. Turning to slide six. I will now turn to the overall financial performance and our balance sheet strength going into this environment. Our first half financial performance was resilient, and we delivered on our key strategic objectives, and that includes balanced growth, capital optimization, funding efficiency, cost reduction, and margin stabilization. Our balance sheet is strong, with a resilient capital base, including a sizable GBP 800 million CET1 management buffer and a prudent funding and liquidity position.
Our lending portfolio is also defensively positioned, and we are not exposed to the most impacted sectors such as travel, high street retail, energy, or commercial real estate. Ian will explain the details of our portfolios later in the presentation. In terms of the balance sheet mix, we achieved good growth in our target areas of business and personal, both up around 6%, whilst in mortgages, we maintained our pricing discipline in the first half. On deposits, we saw continued growth of around 4% in our relationship deposits, which was also very pleasing. As a result, we're able to deliver on our NIM guidance with stability in the NIM at Q1 and a slight expansion in Q2, leading to a first half NIM of 1.63%. Our efficiency drive continued in half one, with costs down 3% year-on-year and a cost income ratio of 57%.
Our transformation program delivered a further GBP 23 million of net run rate cost savings in the period, with a total of GBP 76 million delivered to date. Asset quality remained robust pre-COVID, with a 23 basis points cost of risk. However, given the economic implications of COVID, the outlook for credit impairments has changed, and we have determined that an additional GBP 164 million of COVID-related impairment provisions are required at this time. This will leave us with nearly GBP 550 million of on-balance sheet provision reserves, and Ian will talk you through the detail of this shortly. Finally, exceptional costs were in line with expectations, and we took no further PPI conduct provisions in the period. Given that the PPI complaint uphold rate is tracking lower than expected, we anticipate that the current run rate could lead to a provision surplus.
The combined outcome is that we delivered an underlying profit of GBP 120 million with a 4.6% underlying return on tangible equity. The statutory profit after tax of GBP 22 million and retained a resilient capital ratio of 13%. I will now hand over to Ian to take us through our balance sheet strength and the financial results in detail, and will return to discuss our thoughts on the full year and beyond.
Thanks, David.
Ian.
Good morning to those of you in Europe, and good evening to those dialing in from Australia. I hope you're all keeping safe and well in these difficult times. I'm going to spend a bit of time now, as David said, talking in detail about the numbers. There's a lot of ground to cover, but I'll try and keep it brisk. Pre-COVID, we delivered a robust set of results, and it was encouraging to see the business perform well, particularly on net interest margin and cost delivery. Of course, that's an important backdrop to the current environment, and it underpins the confidence we have in our strategy and our ability to execute. Before I cover the results, I want to talk about our balance sheet, so our loan book, funding, and capital.
The headlines will be familiar from our investor call on 31 March, I do want to give a little more detail now and make it clear that the bank is well prepared for whatever may lie ahead. On 31 March, I made the point that not all portfolios are created equal. Today I'm going to dig into our loan portfolios, walk through how we thought about the risks that lie ahead, and how that's helped us construct our ECL provision at 31 March. I'll talk about our capital position and how RWAs could evolve over the next 18 months. Finally, a few words about funding and liquidity. While the macro environment is clearly uncertain, we believe the way we've built our portfolios and their natural defensive nature provides a degree of resilience to the aftermath of this economic and social shock.
Our lending comprises prime resi and buy-to-let mortgages. A carefully constructed relationship-driven business portfolio with a clear sector focus, and that's helped us avoid unhelpful concentrations. A high-quality personal lending book predominantly focused on the more affluent customer credit card portfolio. Consequently, quite frankly, the margin strength of portfolio reflects the lower risk inherent in our lending. Pre-COVID, the portfolios performed robustly and in line with expectations. The business as usual total cost of risk was 23 basis points, with a low incidence of specific provisions in the business portfolio. Since the onset of the U.K. lockdown, and in line with our peers, we've seen payment holiday requests in mortgages and personal, and requests for capital repayment holidays in business.
Now, while I sort of hesitate to borrow the phrase du jour, I think we're past the peak because the number of requests has slowed markedly in the last couple of weeks. We're comprehensively engaged in supporting our business customers where appropriate, via the range of government guaranteed loan schemes. As David says, we've booked a COVID-19 related impairment provision on top of the normal ECL of GBP 164 million, based on scenarios modeled with conservative economic assumptions and supplemented by expert judgment. This took our on-balance sheet ECL provisions to almost GBP 550 million as at the end of March. It's a mixture of judgment and rigorous analysis, art and science, if you like, but we believe that this provision is prudent. It reflects the significance of the economic shock, but also the potential shape of the economic cycle.
The modeled demand and cash flow shocks for businesses, and the risk of elevated unemployment means that additional provisions are predominantly related to the business and personal portfolios, with limited impact on our mortgage book. There's limited CET1 impact from this additional provision, as around GBP 90 million has been set off against our excess expected losses capital deduction, and the remainder benefits from IFRS 9 transitional relief at 85%, leaving us with a resilient CET1 ratio of 13% and a GBP 800 million buffer to the regulatory minimum. Turning to Slide 8. I'll talk through the portfolios in more detail, but let me say up front that our key credit metrics currently remain robust across the board. Unlike some of the peer group, we saw hardly any specific provisions triggered by COVID-19 as at 31 March .
Our mortgage book is primarily owner occupied with low average LTV and LTIs, and it's been built over the last few years under the MMR regime, adhering to prudent underwriting standards, and there's nothing niche or legacy in that book. The business portfolio is predominantly lending to larger and mid-market SMEs, typically in lower impacted sectors. We have very limited exposure to sectors in the front line of economic damage, and two-thirds of the portfolio is fully or partially secured. 80% of our personal portfolio is the prime affluent credit card book that was built on the lessons learned in the GFC downturn. On slide nine, mortgages. 76% of the book is owner occupied, so we've got no subprime, no self-serve or other specialist categories. Portfolio arrears stand at 0.4%, which is nearly half the industry average.
Owner occupied has got strong credit metrics, low LTV, with 21% of the book being greater than 75% LTV and 2% greater than 90% LTV. Good affordability with average loan to income of three times. Critically, the portion of the book in the danger zone of high LTV and high loan to income is negligible. Our buy to let business is focused on amateur landlords with small portfolios and high levels of income to ensure resilience under stress. This is a strong portfolio. The key dynamic is the length of lockdown and the effectiveness of the government support measures, and we expect the latter to be very helpful in supporting incomes for customers. It's still too early to determine exactly how much of an offset that provides. At around about 15% of our customers have been granted payment holidays, which is in line with the peer group.
Slide 10, our business portfolio. Look, there's no question that this is a tough time for businesses. We believe that our portfolio is well constructed, but this stress, of course, is like no other. We think it's helpful to look at our portfolio through two lenses, sector exposures and the size of business customer. On a sector view, we think of the book in four main buckets, and almost 80% of our lending is in sectors that we expect to be less impacted than others, and less than 10% of the book is in the highest impacted category. More than ever, cash flow is critical, along with access to alternative sources of funding. It's helpful that our book comprises stronger cash flow mid-market SME businesses, typically with turnover of GBP 2 million to GBP 100 million, accounting for 96% of lending by value.
These larger SMEs are usually more established with higher numbers of employees, cash reserves, and access to advisors. Unprecedented levels of government support are on offer with rate reliefs, wage support, and cash grants, all helping alleviate the pressure. We expect almost all of our business customer base will be eligible to benefit from one of the government schemes. Finally, our entire business portfolio is relationship managed by dedicated RMs of long standing, and this is critical. It means we've got around 300 expert RMs on the ground proactively contacting customers, keeping right up to date with the customer, providing support, and helping to manage emerging risks. On slide 11, I thought it'd be helpful just to show you how we think about the book in those four buckets I mentioned.
55% of the book is in sectors that we think are least exposed to damage from the COVID-19 situation. Agriculture is our largest concentration. We're very well diversified and collateralized across different subsets of agri, early indicators show limited impact of the crisis on farming revenues. One tail risk is the availability of migrant labor over the summer, overall, our book remains well positioned. In health and social housing, our principal exposure is care homes and specialist care. Revenues are resilient as beds are very much needed in this crisis, social housing hasn't seen much impact so far. Revenues are holding up, supporting book quality. 22% of the book is in lower impacted sectors. This includes specialist hotels, primarily city center, around two-thirds in central London.
While these are currently impacted, the strength of the location and the brands mean we are positive about their future relevance and ability to trade profitably, and exposures are underpinned by good levels of collateral. As you know, we've had limited exposure post GFC to commercial real estate, and what's on our books is low LTV and very conservative covenants, supported by strong cash flows and coverage. We have zero speculative development CRE. Manufacturing is well diversified, and most of our customers can flex costs by furloughing staff, reducing variable costs materially, effectively hibernating the business. Around 14% of our business lending is to sectors that are more immediately exposed in this stress. The degree of impact depends on the type of business supported, so supermarket service providers are less impacted, while entertainment firms more so.
We have private equity backing in this space, and often that provides an additional layer of support, and there is also invoice and asset finance available. We think of the last bucket as higher impacted sectors, and while we're not in high street retail, we do have some exposure to small retail businesses where flexibility and security can be more limited. In the entertainment sector, our exposure is typically to larger businesses where the loan is syndicated across multiple lenders. Crucially, we don't have meaningful exposure to the most impacted sectors such as airlines, oil and gas, travel, high street retail, or speculative development CRE. Turning to slide 12. We disclose GBP 3.6 billion of undrawn exposures in our business portfolio, but it's worth noting that more than half of that balance is unlikely to be drawn.
Revolving credit facilities and overdrafts are the real areas where we might expect to see drawings. We haven't seen much sustained drawdowns to date, and certainly not to the extent that I've read about in other banks. The important thing to understand, though, in terms of RWA impact, is that we already hold circa 70% of the potential drawn RWAs against these lines. Not a huge capital impact should the customer decide to draw down. Finally, looking at what's been happening with the customer base since the middle of March. As David said, we helped around 4,500 business customers manage their cash flow via lending facilities, overdraft extensions and increases, capital repayment holidays, and through the government-backed loan schemes. Moving now to slide 13 and personal.
In our personal portfolios, experience shows that unemployment is the highest driver of impairment, and that certainly is what drives our impairment models. Our unsecured segments comprise prime affluent customers, so we start from a good place. Our cards portfolio has been deliberately built more conservatively than the market average. The GBP 4.2 billion book is 90% a heritage Virgin Money portfolio, largely built up since 2015, and the heritage CYBG portion primarily comprises PCA customers who are known to the bank. We targeted low risk demographics with low levels of indebtedness and robust underwriting standards. Average age, average income, and the proportion of homeowners is high. With low exposure to self-employed customers and with debt-to-income and persistent debt much better than the industry average. Personal loans are a mixture of long-standing banking customers and more recent digital customer acquisition in good quality keenly priced business.
Again, with lower debt to income, higher affluence, and lower self-employed. Moving on to slide 14. I just want to say a few words just about the cards portfolio. The cards management team has got deep experience. They managed scale credit card portfolios through the GFC. On this slide, I wanted to explain just how that experience has been applied in building and managing our book. Now, stress losses can be material in unsecured lending. However, we know that on top of the standard risk metrics, customer affluence makes a big difference in mitigating losses. The team have used that experience to construct a portfolio skewed towards higher affluence. Building affluence predictors into our underwriting with higher score cutoffs, no acceptance for customers with higher indebtedness, and a strict policy of no downselling.
Third-party benchmarking shows that VMUK has achieved higher quality originations every year relative to industry averages. We've also focused the book on a lower risk product set. Market-wide balance transfer arrears have been better in stress and through the cycle. Our portfolio was built BT heavy, with lower debt service costs for the customer. Only a small proportion of customers move post-promo in the next six months, limiting near-term risk. On to slide 15. As a result of that careful portfolio construction, our vintages perform much better than the industry. Typically, our cards delinquency has been one-third of industry levels. As I mentioned, unemployment is the key driver of stress losses. While we won't be immune, our affluent customer profile typically has larger resources to fall back on in stress.
Now, government support on income is again helpful, and VMUK is focused on delivering the right support for customers over and above what's required of us. So far in our personal business, we've had a low level of requests for payment holidays, less than 2% in cards and less than 6% in PLs. Again, we're seeing the rate of applications slow markedly over the last couple of weeks. That's our loan portfolios, and I'll just spend a bit of time now talking about how we've tackled impairment provisions. On slide 16, I'm going to focus on how we've developed the COVID-19 impairment provision. Firstly, we've seen next to no COVID-related specific provisions in our business book to date. The impairment provision we have taken is forward-looking and has been developed using a combination of economic scenarios and expert judgment.
The art and science I talked about earlier. I should also just be clear to listeners that we source the economic forecast from Oxford Economics. We don't develop our own economic forecasting. Much has been said in relation to the shape of the shock and the subsequent recovery. Our core belief is that no matter the scale of the short term disruption, the U.K. faces a long period of low growth, declining asset values, and structurally higher unemployment. Consequently, we weighted the economic scenarios in our IFRS 9 models 100% to our pre COVID-19 downside scenario. We show the most important of the economic variables in this scenario with the purple lines on the graphs on this slide. It's this economic modeling that delivered, I guess, the base provision for the COVID ECL.
Next, we also reviewed each of our portfolios, looking at specific vulnerabilities, and we increased the base provision for those items. We assumed a proportion of the customer base on payment holidays will move into genuine forbearance and arrears. We looked at elements of the business portfolio that might suffer under stress, and we allowed for adverse ratings migration on the more exposed segments of the book. Finally, the third layer is really the conscious that more recent economic forecasts have assumed a deeper economic shock than was anticipated a few weeks ago. We also modeled the impact of a pandemic scenario developed by Oxford Economics on our business and cards books. The key attributes of this scenario are a deep 2020 GDP reversal and unemployment spike into Q1 2021, followed by recovery in 2021 and beyond. Those are the red lines on our graphs.
This last piece led to a further top up of the provision. Now, we believe that the combination of these three elements leads to an appropriately conservative assessment of ECL due to COVID-19. The key drivers of credit losses are the duration of lockdown, the pace of economic recovery, the resulting unemployment rate, and the effectiveness of the government support schemes. Taking all of that together and turning on to slide 17, our aggregate COVID-19 provisions are GBP 164 million, and that's made up of GBP 146 million charged to the income statement and GBP 18 million of existing economic uncertainty overlay. The COVID-19 provision has increased our ECL by nearly 50%, and we think that is some way more than our peer group. We now have GBP 542 million of ECL provision reserves, providing good coverage of our portfolios.
Most of the increase in the stock of provisions is against our business and personal books. Our CET1 ratio was largely undisturbed by the additional impairment charges. GBP 90 million was absorbed by our EEL deduction or in Aussie parlance, regulatory expected loss in excess of eligible provisions. The remainder attracted IFRS 9 transitional relief at 85%. Now, we're not giving guidance on full year cost of risk. However, given that our start point is prudent and based on the current economic outlook for 2020 and 2021, we wouldn't ordinarily expect further provisions of this magnitude. However, the forward-looking picture is highly uncertain, and there's obviously risk of further deterioration. I wanted to say a few words about RWA migration and slow. We deal with that on slide 18.
I really wanted to explain why we're less exposed to RWA migration risk in the short term, particularly during the second half of this financial year. I also want to highlight some work we have underway that, subject to regulatory approval, delivers material RWA benefits. We have no capital markets exposed business. Our counterparty credit and CVA risk-weighted assets are very low, so we have negligible exposure to RWAs that typically move most quickly and materially in volatile markets. Currently, our personal portfolios are assessed for capital requirements under the standardized approach, and RWA densities are fixed at 75% with low risk of inflation due to environmental factors. Our business banking portfolio is primarily on foundation IRB, which is less sensitive to changes in external conditions. The models use a fixed regulatory LGD, and so RWA migration is only driven by PD fluctuations.
Our mortgage books are on advanced A-IRB. We will see RWA inflation if arrears drive higher PDs. Of course, house price index reductions can affect LGDs. Our RWA density is already higher than peers, reflecting model conservatism, and we upgraded our LGD models in the half just ended. Importantly, in our models, HPI reductions don't start to bite until we exceed 20%. We're also progressing several RWA initiatives that we estimate will, in aggregate, reduce RWAs by 5%-10%. Firstly, as you know, we're transitioning our cards portfolio to IRB. The application is in with the PRA, and on the current timetable, we should receive accreditation by the end of FY 2020. In addition, we have several initiatives in our business RWA model suite.
For example, we just applied to move from a contractual maturity approach to one based on effective or behavioral maturities, that change in tenor reduces the RWAs. These initiatives will also yield savings over the next few quarters. Taken together, the scale of the cards and business initiatives in the second half of FY 2020 is similar to the model related RWA increases we saw in the first half. Finally, we are pursuing a material improvement in mortgage risk weights by moving our heritage models to the hybrid approach. This is a key industry initiative that by and large, was expected to lead to net RWA increases for many of our peers, whereas we had always anticipated it would reduce RWAs for Virgin Money UK. That one's a longer project.
We're working on the basis that we'll secure approval for the new models in the first half of calendar 2021. All of these developments are, of course, subject to PRA approval, and the impact is dependent on the portfolio at the time and any conservatism the PRA would like us to add. However, we have a good track record of making our case. Remember, in 2018, we were the first bank to secure IRB accreditation since, I think, 2007. Turning to slide 19. Our CET1 ratio was resilient at 13%, equivalent to a management buffer of around GBP 800 million against regulatory requirements of 9.9% after the reduction in the countercyclical buffer to zero. In addition to the substantial RWA improvement opportunities we talked about just now, we're managing our capital carefully in the current environment. Just a few things for you to note.
We will moderate our growth in higher RWA density lending, such as unsecured in the short term. In business, we're focused on supporting existing customers, and we don't expect to see much new customer acquisition in the next couple of quarters. PPI is well under control, and I'll talk about that a wee bit later. In addition, we're delaying all non-mandatory spending. Our integration and transformation programs have been paused. The Virgin Money brand relaunch and heritage rebranding are also paused. We wouldn't get the right return on investment in the current environment, but we expect these to be rephased into FY 2021. That all leads to lower integration and transformation costs in FY 2020. We will expect to incur these costs in the future as we gear back up. Finally, turning to slide 20. Funding and liquidity are in good shape.
Only around 12% of our funding is wholesale, and the deposit franchise is ticking along very well. We started the year with very strong liquidity as a precaution heading into the FSMA Part VII, and the LCR at 139% remained very healthy as at 31 March. On the non-deposit funding side, GBP 7.1 billion of TFS will be refinanced into TFSME. The bulk of our wholesale funding is more than one year in maturity, and our funding plan always allows for nine to 12 months of market shutouts without government support. Since 31 March, we've seen substantial deposit increases because customers have moved to conserve cash. We had GBP 1.2 billion of deposit growth in the month of April, with lending flat to slightly down.
Some of this deposit increases we expect to be temporary, and I guess the only impact in the short term is likely to be some mild NIM dilution. That's the sort of roundup of the balance sheet. Turning to slide 21. I hope you found that detail helpful. I'm going to move on to discuss the key points in the first half results, and then I'll hand back to David to close. Turning to slide 22. Our underlying performance in the half was resilient against a tough backdrop. The income environment remained difficult, we're down to 3% year-on-year due to NIM compression. Although we improved 3% half-on-half with the margin expanding to 163 in the second quarter and higher other income from sale of gilts.
Costs fell 3% year- on- year, offsetting the income pressure, leaving our cost income ratio stable at 57%. In pre-COVID, our credit performance was strong with a cost of risk of 23 basis points, including COVID, our cost of risk was 63 basis points with the COVID impact annualized in that calculation. Underlying PBT at GBP 120 million was lower year on year and half on half, driven by the COVID-19 impairment charge. Our underlying return on tangible equity was similarly impacted down circa 6 percentage points to 4.6%. Although prior to COVID impairment charge, we delivered a double-digit return. Turning to slide 23. We returned a statutory profit after tax of GBP 22 million. Exceptional charges of GBP 127 million in the first half had two main drivers.
Integration and transformation costs at GBP 61 million. GBP 57 million for acquisition accounting unwinds. There were no legacy conduct charges in the half. We had a tax credit of GBP 29 million due to the revaluation of deferred tax assets following the announcement by the government that it would maintain the corporation tax at 19%. TNAV improved GBP 0.033 half on half, helped by an increase in the defined benefit pension asset. Turning to slide 24. We continued to do well in deposits, improving the mix by focusing on our relationship deposits in accordance with our strategy. Deposit volumes eased in Q2 as we took our foot off the gas on savings and TDs following a strong first quarter. Relationship deposits grew strongly, though, up 4.3%. We continued to capture a leading share of RBS switching deposits with GBP 177 million in the half.
We saw a modest improvement in the cost of deposits. We did some back book repricing in March, those reductions will feed through in due course. Of course, we'll expect to see further price reductions as a result of the base changes that we saw during the month of March as well. Finally, wholesale funding costs increased by 9 basis points as a result of a richer mix. On slide 25, a few words on lending growth. Our growth and mix was in line with our strategy. Mortgages were down 1% on September 2019, but broadly flat compared to the first quarter as we continued to optimize for margin management. Business lending grew 5.7%, with business as usual, boosted by a contribution of GBP 127 million from RBS switching. Lending growth in personal was driven by cards, where balances were 6% higher half on half.
It's hard to call what the second half of FY20 might look like for loan growth. Our best guess is that it'll be more muted than the first half. Really the components play out like this. The curtailment of mortgage market following lockdown means we'll see negligible new business, although it will be partially offset by higher retentions. We'll see facility drawdowns and government support scheme lending replace BAU origination in business. We're unlikely to see the growth rates we had in H1. We will be throttling back personal lending to some extent. On page 26, speaking about margins. Book yields were encouraging in this half. Yield compression abated a little in mortgages as we flagged at our results last year. Business yields were trimmed really due to lower LIBOR rates.
Personal saw a good increase as a result of better pricing, better mix, and also continued seasoning of the cards book with a stronger EIR outturn relative to our prudent assumptions. The improvement in the group's net interest margin in the half followed the shape we set out at Capital Markets Day last year. A richer mix of lending, better deposit costs from our focus on relationship customers, and a wholesale funding headwind from refinancing TFS and building MREL. We also benefited from starting to run off some of the excess liquidity we were holding for FSMA Part VII, and that, together with back book deposit repricing, is one of the key contributors to the improvement between Q1 and Q2.
The 65 point base rate cut in March hits our rate sensitive assets almost immediately, with a bit of a lag in repricing deposits because of the customer notice requirements. What that means is we'll see an immediate step down in our Q3 NIM as the yields on our SVR mortgage portfolio, liquid assets, and elements of the business book reduce, with some offsets from deposit repricing in TFS. We get a good chunk of that reduction back in Q4 as we see deposit repricing catch up the lag to assets. Our exit rate into FY 2021 is a bit stronger. On that basis, we now expect FY 2020 to finish squarely in the 155-160 basis points range.
The outlook for net interest income will obviously depend on volumes, which as I mentioned, are expected to be a little more subdued in the second half. Turning to slide 27. The shape of OOI is as we described at the November results last year, after stripping out the benefit of a one-off sale of gilts. Dealing with that upfront is not a regular practice for us. In March, we sold a small chunk of our holdings to take advantage of attractive market pricing and to lock in a gain of GBP 16 million. We're reinvesting the proceeds into other high-quality liquid assets. Core divisional non-interest income is broadly flat on H2 2019 and down slightly on H1 2019. Mortgage and business fees were both solid.
Fee income in personal was down, largely due to the implementation of the new rules on overdrafts and returned item fees that we flagged last year. Credit card income was also down slightly in March with some COVID impact on card spend. Finally, as we flagged before, investments income was negligible in H1 as the ASI joint venture is getting up and running, and we don't expect material JV profits in the near term. On slide 28, a few words on costs. Our costs were down GBP 15 million, 3% year-on-year, and broadly flat half-on-half.
Last November, you'll recall we guided to a net reduction of costs after deducting GBP 11 million for the transfer into the ASI joint venture of more than GBP 30 million, and that this would take the cost base down below 900 million and was predicated on the delivery of a further GBP 50 million of run rate synergies from integration and transformation. On 26th February this year, we announced significant restructuring proposals that would lead to the closure of around 52 branches and around 500 colleague exits. Following the escalation of the COVID situation in March, the board decided it would be in the best interests of customers and colleagues to temporarily suspend the restructure until further notice.
That means deferring the benefits from those initiatives with a consequential impact on the cost outturn for the year. We now estimate that underlying operating expenses for FY 2020 will be GBP 20 million higher than originally guided. A number of initiatives were delivered as planned in the first half, increasing cumulative run rate net cost savings to GBP 76 million. Key cost saving initiatives included 400 role reductions and consolidation of office space. Restructuring and rebranding costs in the period were GBP 61 million, lower than planned, and commensurate with the deferral of initiatives. Our overall integration and transformation plans remain intact, including the target for GBP 200 million of net cost savings and the total cost to do that we committed to, so still heading to GBP 780 million.
However, COVID has introduced a delay, the delivery profile of those benefits and costs will be rephased once we know how the U.K. will return to work after lockdown. On slide 29, briefly on impairment. Clearly the focus for impairment is on the COVID ECL requirement, which I've talked about in detail. In business as usual, portfolio continues to perform very well with a pre-COVID impairment charge of GBP 86 million or 22 basis points in line with expectations. The small increase on last year is mainly about the richer mix of the book. Really the only item to call out here is the improvement in gross cost of risk in personal, which fell 61 basis points half-on - half.
This was driven by a one-off model recalibration in H2 last year, also some dilution effect of growth in good quality assets, all provisioned in stage 1. Our principal asset quality metrics are solid March 2020 versus September 2019, that continued into April and reflects our focus on responsible credit decisions and our control of the risk appetite. On page 30, a few words on PPI. We've made really good progress on PPI, processing IRs and complaints. GBP 161 million of the provision we used in H1 2020, the remaining provision covers the assessment and finalizing remediation on the complaints from the IR population, plus some other elements of tidy up, such as the official receiver discussions. At the end of September 2019, we had 50,000 complaints and around 325,000 IRs.
We have closed all of those complaints and have processed all but 8,000 of the IRs. The IR to complaint conversion rate was slightly higher than expected. We ended up with around 100,000 complaints from the IRs, and we processed 25,000 of those, with 75,000 still to do. More importantly, the uphold rate on the complaints has been much lower at around 25% over the past six months versus 40% provided for. If that uphold rate continues and nothing else comes out of the woodwork, it implies a potential provision surplus. However, we won't count our chickens just yet. Processing is paused for COVID-19. We're about to restart, and it's possible the hiatus may mean that closing out PPI takes a little longer than we anticipated, but we will be done and dusted later this year. Turning to slide 31.
We had in the first half net underlying capital generation of 24 basis points. After absorption by exceptional and non-operating items, the CET1 ratio was around 30 basis points lower at the end of the period. A couple of things to call out here. As explained earlier, the CET1 ratio has not been affected by the increased ECL provisions for COVID-19. RWAs absorbed almost 60 basis points to CET1 in the period. Half of that related to a scheduled upgrade of LGDs in our heritage VM mortgage models. As I flagged earlier, we anticipate a broadly equal and opposite impact from RWA model initiatives in the second half. The 22 basis points of other primarily reflects the group's pension scheme contributions paid in fiscal Q1. We've agreed with the trustee to suspend pensions deficit contributions in calendar 2020 as our triennial valuation is underway.
Other also includes some negative reserve movements. Now, we flagged consistently that FY 2020 was the tightest year for capital generation and that we expected to see a stronger print in FY 2021 and 2022 as cost savings kick in and restructuring and acquisition unwinds abate. Despite the COVID-19 pause, that remains our overall view. Total capital at 19.5%, and UK leverage ratio of 4.9% remain robust. At 25.6%, we remain comfortably ahead of our interim MREL requirement of 18%. The final MREL compliance deadline of 1 January 2022 is one of the few regulatory initiatives that has not been relaxed. We had our next MREL trade oven-ready in mid-February, just as markets went south. We'll look to resume MREL build later this year. We've got about GBP 1.5 billion to go. Finally, turning to slide 32, and I'll look forward.
It's highly uncertain how the U.K. economy will emerge from lockdown. It'll take some time to see the effects of policymaker actions and ultimate impacts on the real economy and our customers. We believe we've got a good line of sight to the likely outturn on NIM and costs for FY20. We're happy to guide on those. We expect NIM will come in between 155 and 160 basis points for the full year, and we will aim to manage underlying operating costs lower than GBP 920 million, including the bank levy. Over time, the board aims to return to paying dividends and build those sustainably. We'll take stock closer to the year-end on dividends for FY20. Consensus expectations of low to nil for FY20 make sense for now.
Medium term, we still plan to improve our returns through the three key financial drivers we set out at Capital Markets Day last year. A richer mix of lending, lower deposit costs, and GBP 200 million of cost out. Nothing has changed in that regard. The U.K. lockdown means we'll need to revisit and rephase our delivery plans. David will give a glimpse of how we're thinking about that. Realistically, it'll be later this year before we confirm up on those. Thank you for listening. I realize it's been a long session, but I hope you found it helpful, and I'll now hand back to David.
Thanks, Ian. I will now close out the presentation by touching briefly on our strategy and the opportunities that may arise based on what we're learning in this environment. If we turn to slide 34. While it feels like a long time ago now, the purpose, ambition, and strategic priorities that we outlined at our Capital Markets Day back in June remain valid for the future environment. The targets we set for 2022 also feel appropriate for us to aspire to in the medium term. As Ian has referenced, the timeline for delivery is now more uncertain until the impact of COVID-19 becomes more visible. In the short term, though, we need to focus on supporting our customers in difficulty and protecting our capital, and we've therefore taken the decision to defer a number of the transformation and rebranding programs that were scheduled for the second half of 2020.
We see this as a temporary delay and plan to continue with these in due course. We have, however, learned a huge amount over the past few weeks about our capacity to design, build, and deploy projects and propositions at scale in a fraction of the time we ever thought possible. As we continue to learn, I expect that we will be able to develop more efficient delivery models for the post-COVID-19 world. If I give you a couple of practical insights, we were able to stand up the three new government support schemes within two weeks with a digital application process. We've also enabled the entire bank to work with a cloud-based Microsoft suite of products, and that was done in three weeks. These would have taken us a lot longer in a pre-COVID world.
The rapid adoption of digital solutions as a consequence of COVID-19 presents Virgin Money with new opportunities in the future. We are a smaller, more agile bank than some of our competitors, and our strong digital capabilities should allow us to leverage the industry-shaping forces that COVID has unleashed. We expect to be able to accelerate our existing plan to fully digitize our bank as soon as the environment stabilizes. The board and my leadership team are looking closely at the changing business models, customer behaviors, and employee expectations, and are developing an understanding of the opportunities that could emerge from this rapidly evolving environment. We fundamentally believe that an agile, purpose-led, full-service digital bank can innovate at speed and will be able to create significant in the market. We will also support [audio distortion]
[audio distortion] You're just cutting out a bit there, David. It's breaking up a bit, I'm afraid. Not sure if you can improve your signal.
Can you hear me now?
That's a bit better. No, it's still really cutting out, I'm afraid. Still gone, eh, David?
Can you hear me now?
Yes. Perfect.
Okay. Sorry for the interruption. You should see [audio distortion] in developing an understanding of the strategic opportunities that could emerge from the rapidly changing environment. I believe that an agile, purpose-led, full-service digital bank can lead to significant new opportunities in a digital marketplace. We will support our customers as they update their business by implementing flexible digital solutions. [audio distortion]
Sorry, David. It's gone again, I'm afraid.
Okay.
David, while you struggle with that, let me pick up.
Yeah. I've got it now.
Yeah?
Yeah.
David and the board and the leadership team are looking at changing the business models, responding to different customer behaviors and employee expectations. We think that being agile and full-service digital will allow us to innovate at speed and should create more opportunities for us. We're going to support customers as they update their business models, and ways of working, through just flexibility and digital-enabled mobile workforce. We're seeing material shifts in customer behaviors with digital enrollment up 40% in April. I think that plays well to our digital ambition. The combination of those factors should allow us to deliver a better customer experience, but also more efficiently, with greater convenience to customers and greater flexibility to colleagues. We think we're going to learn some important lessons and take some opportunities from the things we've done in this crisis.
In summary, we're confident that strategic priorities are right for the medium term. In the short term, we're going to focus on supporting customers and protecting capital. I guess that's the key priority for us, certainly during this difficult period. Finally, just turning to slide 35. Reminding everybody we've got a defensive balance sheet. There's nearly GBP 550 million of on-balance sheet provisions to absorb future losses. A disciplined approach to risk management, as demonstrated by the regulator's willingness to accredit us as an IRB bank. Our capital base is resilient with a CET1 ratio of 13%, incorporating a management buffer around GBP 800 million. We have several opportunities to further improve capital resilience, as I talked about earlier. Allied to all this is the unprecedented level of policymaker support across both the government and regulators.
While these are clearly not a panacea, the scale of support on offer should help mitigate some of the worst impacts from the pandemic. Finally, we've learned that you build a franchise in difficult times based on how you treat your customers. I'm confident that the support we're providing customers, colleagues, and communities will ensure that we come out of this with our reputation enhanced and a business that's built to thrive in the new operating environment. That's the end of prepared remarks. We'll now turn to Q&A. Back to you, Maxine, to see if we have any questions on the conference call now?
If you would like to ask a question, please press star followed by one on your telephone keypad now. We have a question from Alvaro Serrano from Morgan Stanley. Your line is now open.
Good morning. I had a question on provisions and then another one on margins. Can I just confirm the provision taken is equivalent to the 100% pandemic scenario? I know you've taken three different sort of looks at it, but I just wanted to confirm that if we go into 100% pandemic scenario, would there be any additional top-ups? In payment holidays, you also mentioned that you've taken an assumption of some of these going into non-performance. Can you give us details on how many of your customers have asked for payment holidays and how many of these are you assuming will go into non-performing? The last one on just on margin. Your cost of deposits at 98 basis points. What have you assumed in your margin guidance that that comes down to by the end of the year? Thank you.
Okay. Thanks, Alvaro. I'll take those numbers-related ones. The pandemic scenario, the pandemic shock, was applied to our business and cards portfolios, so not across the whole book. It's important to understand the way we build this up is the sort of significantly lower for longer economics applies across the whole portfolio. That's the base. It was really an additional stress to take account of a sort of major short-term spike in the pandemic, and that was applied to business and cards. It actually resulted in a relatively small top-up to the overall piece. Our core economic assumptions, I think, are suitably prudent. Your second one on cost of deposits. We're not precisely specifying how far that'll go. There's a reasonable opportunity, both in terms of just what happens on those that automatically reprice, but also in our savings book where we have admin rates that we can move according to market. We've seen those come down from highs of sort of 160 towards, on our savings portfolio, down towards 100 basis points.
On payment holidays, I don't know if you can give any detail there.
I beg your pardon. Yes, you did ask about that. The absolute statistics on payment holidays is in mortgages, 60,000. That's 15% of the customer base, in line with the industry really. We've seen 32,000 holidays in credit cards. That's less than 2% of the customer base. 8,000 holidays in personal loans. That's 6% of the customer base. What we did with those, it's impossible to tell for any banks at the moment, how much of that will translate into genuine forbearance. When we looked at the overlay requirement, we did some sample testing, in order to determine how many of the customers were simply being conservative versus those that might be experiencing genuine distress. We allowed for those in the development of the provision. It's pretty small in mortgages, very small in mortgages. Roundabout we allowed for something of the order of 25% in the personal portfolio.
Thank you.
We have a question from Robin Down from HSBC. Your line is now open.
Hi. Perhaps this one is aimed perhaps a little bit more at David Bennett, as he said. Many of us remember sitting through a number of these presentations back in kind of 2007, 2008. Every bank at that point stood up and said, "Our book is better than the average." Yet we had a lot of sort of negative surprises come through. I'm slightly surprised when I look at the provisioning levels that you've got here, that you're applying a pandemic scenario. Where you're assuming that one in 10 people lose their jobs, yet the provisioning that we've got on things like the credit card book, to me feels very light. I think across stage 1 and stage 2 on the personal loan book, you've got provisions coverage of about 3.5%, which just feels low. I wonder whether you're actually comfortable with these level of provisions or whether or not we should just really be expecting another big sort of top-up to come in in Q2 as we actually see the unemployment coming through?
Hi, Robin. It's Ian. I'll take that one. Look, I think keep coming back to understanding the shape of this. Broadly speaking, a sort of a deep GDP shock that then immediately bounces back, doesn't actually feed through to a particular significant increase in impairments in our book beyond what we've allowed for already. The important thing on that sort of peak unemployment of 9.7% is that it recovers quickly, very quickly. I think a combination of we've provisioned on the basis of a more sustained structural increase in unemployment, close to 6% average across our sort of core scenario. The spike simply adds a little bit more to that. I take you back to the mix and quality of our book, and the greater affluence, and therefore the less propensity to incur losses there. That's the basis. Of course, we're comfortable with provisions. We've done a pretty thorough and comprehensive job on those. We think that the quality of our book is the best defense in these circumstances.
Robin, your line is still open.
Thanks. It just feels like in just in pure absolute terms, you just have not set aside enough for the crisis coming through. I guess we'll have to agree to disagree on that.
Robin, I think it'll take some time to sort of settle these down and make some comparisons. I think we've been pretty conservative. Our uplift overall on ECL is we think we've gone further than the other banks across the board. Going back to your point about comfort, we're comfortable. We've done a prudent job on this.
Okay. Thank you.
We have a question from Chris Cant from Autonomous. Your line is now open.
Good morning. Thank you for taking my question. To come back on the same point really, something you said on your investor call at the end of March. We've never seen Virgin go through the Bank of England stress test. At the end of March, you talked about when you run those scenarios internally, you see similar CET1 drawdowns to peer banks, you don't consider yourself to be an outlier. I guess the tone and everything you've said this morning would suggest you do think that you're a bit of an outlier in a good way in terms of both asset quality on the book and the ECL side, also capacity to avoid RWA inflation. How should I reconcile those two sets of comments, one saying you don't think you're an outlier, and then what you've been saying today in terms of how a stress develops. Is it just that the stress scenario you're talking about today is so different from the stress scenarios in the Bank of England stress tests? Thank you.
Hi, Chris. Thanks for that. I think it's always difficult to compare like with like in terms of portfolios and other things. A lot of our read across was to ensure or to test whether we look like a negative outlier, and we don't. We compare comparable portfolios with comparable portfolios. The stress is a bit different. We see this as overall a much more of a sort of lower for longer type stress, in terms of recovery. That is a wee bit different from what was in the ACS last year. I think our sort of CET1 drawdown overall is impacted in those circumstances by ability to rebuild capital generation through cost reductions and other things. We're already taking quite a lot of cost out, et cetera. The comment back on 31 March was more about saying, look, don't assume that somehow we're going to be worse than the other banks. To your point, I think the mix, particularly in the mortgage and cards book, suggests that we do have a stronger, more defensive portfolio.
Okay. Thank you.
We have another question from Ed Henning from CLSA. Your line is now open.
Thank you. Thanks for taking my questions. Can you just touch a little bit on risk weight migration? If the scenario plays out as your forecast, what are you thinking there for risk weight migration to start with?
Hi, Ed. On the scenario that we've outlined, we might expect to see some risk weight migration in mortgages in FY 2021. Simply because of what we'll see there in potential higher PDs and anything that may happen in HPI. The other piece of risk weight migration is one of adverse ratings migration in the business book. That's more sort of split evenly in FY 2020 and the first half of FY 2021. Offsetting that, of course, is the benefits of lower exposures and lower undrawns. That really helps to offset. Net RWA migration isn't expected to be material because of the shape of our portfolio. I hesitate to give numbers on that. We're not really giving more detailed guidance on it, but the shape of it is, we'll probably see a little bit of a net migration in FY 2021, but we're not expecting to see that material on the current base case.
Maybe another way to ask it, if it's not going to be that material, your benefit from the risk-weighted assets rolling off will be more than offset your potential risk weight migration on your scenario at this stage.
It certainly offsets a good chunk of it.
Okay. Just a second question just on your NIM looking forward. As you pull back on cards and also biz growth, a lot of you [audio distortion] with about mixed. That obviously won't play through. Can you just repeat on that?
Ed, you broke up a little bit for me there. Your question was around the impact of lower growth in business and personal on the net interest margin, the mix effect?
Yeah, the mix effect that you've previously called out. Obviously, that now gets pushed out. Is that fair?
I don't think that's going to be hugely different, Ed. As I say, when you combine that with the fact that the mortgage book is probably going to see a bit of net shrinkage certainly in the next six months, just given what's happening in the market. I think we'll be able to maintain the mix progression, albeit on a slightly smaller aggregate loan number. Our own base case consensus for the end of this year is probably a wee bit better than consensus. I think the mix will be preserved.
Okay. Thank you.
We have a question from Guy Stebbings from Exane BNP Paribas. Your line is now open.
Thanks for taking my questions. The first one was just to come back to margin. The step down that you're talking to in the second half of this year is much lower than some of your peers. Presumably, the lower starting and lower structural hedge contributions, that helped. I would have thought the drop in the card balances and measures to support the consumer, such as interest-free overdrafts and things like that, would have had a thinner impact or a thinner headwind into the second half. I don't know if you could talk to your level of conviction on the NIM guidance, given the uncertainty that's out there and also the phasing. I mean, it sounds like you're talking about Q4 NIM being up on Q3, given the deposit repricing. Is that right? What should we think about into 2021?
Sorry, just a quick tag on to the NIM. Is it right to assume that the better yield on the card book has been helping to offset some of those other headwinds? It certainly looks like that happening in Q2. The second question was just on PPI and the provision surplus. You might take a judgment there. If you could help at all in thinking around the sizing of that GBP 280 million, which is currently less. I mean, on the face of it certainly looks like more than a single digit million release if that run rate holds. Any comment would be very useful. Thanks.
Okay. Thanks, Guy. Morning. On NIM, our level of conviction is very high. We have a really good handle on our book, and if you do the arithmetic on the basis of how we've guided in the past, we used to talk about GBP 10 million for a 25-basis point cut. If we assume that 65 is as good as 75, you do the arithmetic on that, you factor in a quarter of lag in deposit pricing, you actually get to a pretty good estimate of what we think the income effect will be in the second half. And to some extent, I had to be restrained by my colleagues from giving you quarterly NIM guidance going forward, mainly on the basis that we do feel pretty confident about it. Why are we different from the others? First, I think it's shape of the book.
In looking at this, clearly the two banks last week that disclosed significant step downs were Lloyds and Barclays. We're much more mortgage heavy than those banks in terms of proportion of the book, and we're also much more fixed rate mortgage heavy compared to that. Our SVR balances are just a shade over 6%. First of all, we don't see the same extent of challenge there. The higher business books from the other banks make for more sensitivity. Certainly in cards, a good chunk of our portfolio is sort of fixed balance transfer on EIR. Again, less sensitive to fluctuations in volumes there. If you add into that a bigger opportunity on deposit repricing, as we talked about when I was just answering Alvaro's question around the opportunity in savings, for example, that the other banks don't have.
The overall start position already having taken quite a lot of the pain, particularly on the mortgage book. I think it's all of those factors that really combine to say we don't suffer as badly as some of our peers as a result of the rate environment. For example, overdrafts is a good one. Our overdraft penetration is much lower than our peers. To some extent, that reflects a sort of previous "weakness" in the book in that we didn't have compelling products and product penetration in overdrafts. The impact of offering a GBP 500 interest-free element to overdrafts for the crisis means that it's easily absorbed within our guidance. The other thing to understand is that our overdraft proposition in this new world was always much more competitive and customer friendly than other banks.
Other banks are charging rates of 39.9% and have had to reduce those in response to the crisis. We went in right from the get-go with a much more competitive 19.9% rate for mainstream overdrafts that we haven't had to touch. I think it is about different positioning, Guy, and also just recognizing that we've taken our lumps on NIM compression in the last couple of years, much more than the other banks have. Structural hedge is the last piece. Our contribution for structural hedge has always been a good deal lower than other banks. We don't benefit from it as much in income terms, and so we don't suffer as much on the downside there. Different positioning. As you can imagine, we've done a lot of soul searching as to why we might be different from what the other banks are talking about at the moment.
That's why we're so high conviction. Your shape is right. Exactly as we guided, the step down in Q3, when deposit pricing catches up, we regain a good chunk of that step down. All other things being equal, we're sort of flat into FY 2021. The benefit from card EIR, really in 2019, we were pretty prudent in terms of our assumptions. We've benefited from that coming into this. That benefit certainly has helped the margin. There's been a bunch of other things there, too. Deposit repricing. Running off a bit of excess liquidity, et cetera. Sorry, I've gone on a lot in that answer, it's really important to convey to you how confident we are in our margin guidance. On PPI, the main thing to convey here is it's going pretty well against our assumption. We're certainly very confident that what we're holding is enough. The extent to which we can get a small release out of that, let's wait and see what happens. There's still a bit of water to flow under the bridge, but certainly at current run rates it would indicate we've got more than we need.
Okay. Very clear. Thank you.
We have a question from Benjamin Toms from RBC Capital Markets. Your line is now open.
Hi both, thank you for taking my questions. You just noted on the presentation the MREL requirement for one of the only regulatory matters measures not to be relaxed as a result of COVID-19. Was that comment one of surprise or a potential flag that there's some expectation that the regulator could do something with the requirements here in terms of relaxing them? Just in relation to your expected excess loss deduction in CET1, have you used all that deduction up now? Just as we go through the next couple of quarters, and as the macro overlay gets allocated to real cash losses, do you have to build that buffer back up again, or does it stay where it is now? Thank you.
Thanks, Benjamin. I think the regulators should have given everyone an extra year on MREL build. We certainly ask for it every week. It isn't something they've been minded to grant up until now. Yes, it would have been nice to be. In terms of the excess expected losses, we've essentially eliminated that deduction from capital. The extent to which it increases going forward, it is about the rate of conversion to actual losses. Sorry, let me stop. It's different from transitional relief. Transitional relief is all about stage 1 and 2, and to the extent that losses are incurred in stage 3, it doesn't attract transitional relief. Excess expected losses are purely about your regulatory expectation. If that exceeds your provision amounts, irrespective of what stage they're in, then you have a deduction. At the moment, the things that are going to drive increased regulatory expected losses, it's probably most sensitive to PDs in the mortgage book. You might see that tick up a little bit in 2021. It's different from transitional relief. I hope that makes sense.
Thank you.
We have a question from Joseph Dickerson from Jefferies. Your line is now open.
Hi. I've got a couple of questions. Firstly, on the unsecured from credit card growth, in particular. The Bank of England data is showing net repayments, which is driving down balances year- over- year in the systems. Are you seeing a similar phenomenon? Effectively the implied growth rate, if there was a higher propensity to revolve, is actually higher than what it looks like optically. Is this current phenomenon something you expect to continue into the second half of the year? That's question number one. Secondly, on slide 30 in your comments that there might be a potential PPI provision surplus. If you could give us some thoughts around the timing at which you might know that. In other words, is it calendar, year-end, et cetera? Basically thinking through the math of it all.
If I look at the operating and processing costs associated with this, there's about GBP 269 million of provision. Is the right way to think of this to apply an uphold rate of 25% versus 40% so you're looking at in the order of magnitude of GBP 100 million kind of surplus firstly, or would you include the processing costs in the calculation? I guess how to try to dimension that somewhat based on the numbers that you've put out today. Thanks.
Sure. Hi, Joe. Thanks for that. On unsecured credit growth, a couple of things and propensity to continue that. I think the first thing is our unsecured growth has also included some growth in personal loans. That has really been about us I suppose growing to more of a natural market share given we've stayed out of that market due to sort of lack of capability for a while. That's been part of our growth. Who knows what will transpire in the next 6- 12 months on that. On credit cards. We've got to sort of look at the shape of our book. The shape of our book is still around 60%-65% balance transfers. Those are not sensitive to general credit card market trends. It's low-cost borrowing for affluent customers, and is largely fixed.
I don't think we're as exposed to short-term reductions in overall credit card balances that the rest of the market would be. Yeah, we might expect to see some lower growth in the second half, but shape of our book means that we're less sensitive. Let me just pause on PPI just for a second here. First of all, there's nothing in our capital guidance or anything like that on PPI. We don't want to be drawn on sort of sizes of potential provision surplus or anything of that nature. It is not at the scale you sort of conjectured there, Joe. I think the key thing to take away from that is the risk of further provisions is very low. Think of it that way.
No threat to capital from PPI legacy conduct is much more important than thinking about there being a sizable release. Just in terms of your sort of high-level calculation, there is remediation and cost to do that apply to all complaints. It is the extent to which, as we process these claims, they turn into claims that are valid for compensation. That is running at a much lower rate, although that rate of compensation, the amount of remediation we pay out, is slightly lower, but more or less in line. Our processing costs are more or less in line. It is not an extrapolation. As I say, on PPI, I think what we should all take away from here is that, a threat to capital from legacy conduct, we think is off the table.
Thank you.
We have a question from Fahed Kunwar from Redburn. Your line is now open.
Hi. Morning, Ian. Just a couple of questions, one on capital and one on kind of loan loss provisions. Just on capital, thinking about the kind of risk weight guidance you've given. Do you think kind of CET1 cross around here, a few of the banks have said that should trough in Q2, not too far can I assume from the current levels. Do you see the kind of 13% and more around the 13% the kind of trough level of CET1, based on the guidance you've given? Just on your MDA as well, any update on the Pillar 2A reduction that you were hoping for as you go through all of the kind of COVID related stuff and still seems quite high. I know a lot of it was for the operational stuff with the CYBG Virgin merger.
The second question I had was just on the provisions. Kind of following up on the earlier questions. Can we get a couple of things? On your kind of 100% pandemic scenario, if you apply it to mortgages, how significant would the loan losses be? It feels like the shape of the curve really is why the credit card losses seem to be a lot less than I think we would have thought if you assume a 10% unemployment rise or an increase to 10%. If you assume a kind of shallower recovery or a kind of less steep gradient on the second leg of your recovery, how sensitive are the loan losses? Do they kind of increase exponentially as we assume that GDP on the second leg of this is lower? Any help on that would be very much appreciated. Thank you.
Sure. Okay. On capital. Yeah, I think certainly in the short term, we'd expect to see a sort of capital low point consequent to our third quarter, so second quarter for the December reporters. The other thing that is helpful for us is the extent to which we see some of those RWA benefits come through towards the end of the year. The help with the build back up. Yes, a sort of trough in the third quarter, if you like. On Pillar 2A, we've got a ICAAP in with the PRA right now, where we think that we've put forward a case for a lower Pillar 2A requirement. We have to wait to see where the PRA comes out on that. It's probably quite an interesting time for them to be thinking about ICAAPs and things at the moment.
We should hear, I guess, sometime early summer on where we come out on that. We might have an update for you in the Q3. In terms of loan loss provisioning, we didn't apply the sort of pandemic scenario to the mortgage book principally because of two things. One is the mortgage portfolio is sensitive to longer term unemployment, so not a short-term spike per se. House price reduction. What we've already baked into our base case scenario, the purple lines on the graph, if you like, on the slide there, is a sort of higher peak to trough movement in HPI and sustained unemployment over the next five years at close to 6%. We didn't think the pandemic scenario was going to move the needle, particularly on mortgages.
Yes. One follow-up just on the kind of shape of the curve on the second leg. How sensitive are credit card and business losses if you assume a kind of a shallower or a weaker recovery coming out of this? Is it exponential or is it quite linear?
Sorry, Fahed. Can you repeat? I didn't get the first part of that question.
Yeah. I just wonder, it feels like your credit card losses are low considering the depth of the recession. Just thinking about how sensitive your scenario modeling is to the second leg. You've got quite a sharp rebound in both unemployment and GDP. How should we think about if that recovery isn't as strong as perhaps you think? How sensitive, what are the sensitivities around that credit card and business banking, but with the losses kind of exponentially increase if the recovery is a lot slower?
Yeah. Okay. It's a good question, Fahed. It is about a sensitivity. It's hard to sort of frame this. I suppose one way of framing it is consensus out there for aggregate two-year losses, FY 2020 and 2021, is a good deal higher than our own view of the world. I guess what you'd have to see is unemployment and HPI to weaken significantly and be sustained across that to get anywhere near that consensus outcome. I guess that's how we think about it.
Thank you very much.
We have a question from Martin Leitgeb from Goldman Sachs. Your line is now open.
Yes, good morning. I was just wondering if you could comment what impact you would expect from the newly launched Bank of England Term Funding Scheme on your business both in terms from an asset and liability perspective, if you like. I was just wondering if that changes anything and how you think about the mix in terms of mortgages versus now business as outlined on the Capital Markets Day, or whether that's essentially principally largely unchanged. Secondly, I was just wondering if you could give us an update on how you're progressing on the Virgin Money current account. Thank you.
Okay. Thanks, Martin. David, just looking for a rest here. If I do the one on TFS, would you pick up Virgin Money current account? On TFS, I guess the way we're thinking about this is to be reasonably circumspect at the moment. What I mean by that is we had an aggressive TFS pay down schedule through the course of this year. Obviously, we sort of paused that. Our GBP 7.1 billion of TFS is not too far away from our initial allowance on TFS. Based on your stock of lending. What that tells us is we'll benefit from the terming out of lower cost funding in a way that we weren't in the previous case. It helps net-net from a margin perspective.
In terms of then how that might drive what you do in terms of business shape and things going forward, I think it's too soon to tell, and it's too soon to tell from two perspectives really. Our ability to draw, and indeed our appetite to draw more TFS will depend on the extent to which then we see net lending growth through this year, because that's the condition to exceed your initial allowance of 10%. We're not banking on that at the moment, just from a prudence perspective. We did see net lending against our 1 January position anyway. In theory, there is some stuff there to be deployed, but we're still just sort of working through that. Then I think we also need to just think about how the sort of competitive landscape plays out.
We weren't a fan in original TFS of pursuing growth, sort of particular growth funded by TFS because you have to pay it back. That refinancing means that it's a sort of temporary benefit, I guess, unless you can refinance it on favorable terms. I think jury's out in terms of just how much that impacts our plans for growth. I think your question was one about how are we thinking about how are we progressing the Virgin Money current account launch. Is that correct?
Yeah. Just more broadly as a kind of update. Obviously, it's one of the key elements of the strategy going forward, having the current account. I was just wondering how this has progressed since.
Sure. Do you want to, David or Fahed? David might be on mute. Okay. Let me take that one while David finds his mute button. The sort of conversion of B to Virgin Money current account, the sort of soft launch of the digital project, I guess, early days, a decent customer response. We're actually selling new current accounts even today. Which surprises me, because I wouldn't have thought that customers would be focused on opening new current accounts at the moment. Digital sales are actually holding up reasonably well. It was always a thing for next year with the big launch of proposition and everything that goes with that. Obviously, as we said in our remarks, there's a bit of a delay as a result of COVID-19. You certainly want to launch this at a time of maximum impact. That'll be sort of early FY 2021 thing.
Our personal deposits, sort of current account deposits have done reasonably well in this first half. I think the other important thing to note is Virgin Money current accounts were about both personal and business, and the business stuff has also been paused. We've seen good core business current account openings and balance generation. Still feel good about it. We still like the early signs from the conversion of B. It's just that things have moved to the right slightly in terms of launch of the big proposition.
Perfect. Thank you. Thank you very much.
We have a question from Aman Rakkar from Barclays. Your line is now open.
Hi, Ian. Just one quick one from me. I'm struggling just a little bit with your cards book, with some of the disclosure on your credit cards book. You've laid out some really helpful stuff about the quality of that book in your slides today, and the prime nature of that book. You also did provide in your 2019 annual report a distribution of balances by probability of default on your personal book. I know it includes some non-card bits and pieces in there. If I look at that, it says that only 28% of your credit cards book has a probability of default below 50 basis points, which is typically the threshold for something that's being considered investment grade or sub-investment grade. You got 72% of your credit card book, which you yourself consider to be sub-investment grade. There's just something about that disclosure, which I'm struggling to reconcile those two views of your card book. Is there something about that annual report IFRS 9 disclosure, which I'm kind of misinterpreting? Thank you.
Aman, I'm going to have to go and do some digging on that. I guess a couple of things. It would be a sort of mixed picture with PLs and overdrafts, not just cards that are in there. We just need to sort through that for you. We can certainly come back to you on that. It's a question of whether these are sort of actual PDs versus regulatory PDs for books that are unstandardized. Again, I don't know if that's playing into it. We'll come back to you on that one.
Okay. Appreciate it. That's a fiddly question. Thank you.
We have a question from Rohith Chandra-Rajan from Bank of America. Your line is now open.
Hi, good morning, and thank you for a very comprehensive presentation and obviously lots of Q&A. I was just wondering if I could come back on provisions and capital, please. The coverage that you show on slide 17, business is obviously a lot stronger than peers, I guess, reflecting the SME bias of the book. Mortgages at 9 basis points, just 20%-30% of what the coverage levels that peers have, and the full 40 on the personal book is maybe 45%-50% of the coverage. On the mortgage book, certainly relative to major banks rather than the specialist lenders, I can't see that there's a particular difference in the mix that you have that would suggest it's materially better credit quality. On the personal book, appreciate the arrears levels that you've shown.
Actually, David Bennett's comments at the beginning reminded me that Alliance & Leicester used to show arrears levels by vintage. I wondered whether you have that comparison for your own book on the basis that actually a lot of your book it tends to be a bit younger and less well seasoned than some of the peers out there. That would be helpful just in terms of trying to compare the coverage levels. Then CET1, just to come back on that, you indicated a reduction in Q3 and then a recovery in your Q4. I note on slide 32, you didn't reiterate the 13% guidance for the full year. I was just wondering if you could help us think about the moving parts for that in terms of transitional relief that you'd expect on further IFRS 9 provisions. The model approvals look like they're potentially about another 30 basis points in the second half and a little bit of RWA inflation. Any additional color on top of what you've already given on that would be really helpful. Thank you.
Thanks, Rohith, and hope we've convinced you on margin there. Look, three really good questions there. I beg to differ on the mix and quality question on mortgages. Maybe just take you back to what did the shape of our mortgage book versus peers look like pre-COVID. What you saw there was we, against a range of the bigger bank peers, have much lower proportion in stage 3. That backs up our half industry average arrears rate, all those sorts of things. A much stronger performing book. What could lie behind that? I suspect it's because if you look at, say, ours versus HSBC, we have followed a similar path in terms of just both the period in which we've built our books and others, and the general shape of it.
We don't have any of the niche specialist stuff that must be driving the different shape of staging in some of the more established peers. The Virgin Money book, back in 2012, was seeded with GBP 15 billion of the very best quality mortgages that Northern Rock had to offer. If we take us back pre-COVID, where nobody was worrying particularly about our mortgage book versus others, I think there are some clear distinctions there. Look at what we've done since then. Since then, we have increased our provisions in the mortgage book proportionately more than any of the other banks. I think we've done a good job, and we've got a really good mortgage book with none of the specialist stuff that used to get done 10 years ago that is still hanging around. It's a book that turns over pretty quickly.
It's been written under MMR. Very comfortable with this mortgage book there. On personal, again, I just wonder if you're being slightly unfair. We talk about the performance of vintages. The arrears performance that we give on the slide that shows that we are about a third of the rest. It's comparing like with like in terms of vintages, so I don't think that use of the book plays into the difference there. It's simply a stronger performance on like for like vintages. On CET1, yeah, I agree. We haven't reiterated guidance at 13% guidance for year-end. I guess you can probably interpret that all things being equal, we'd want to be there or thereabout, just on the moving parts I've spoken about. I suppose 13%, to some extent, we've parked, I think, until we see what lockdown looks like.
What I mean by parked is we sort of go, look, we've got a big stock of capital, a big buffer to MDA. I know Barclays were talking about it in terms of threshold to MDA last week in those terms. I think we're going to just manage that stock of capital the best we can. On our base case, I wouldn't expect us to be too far off our 13% by end of the year, but we will have to see what transpires. There are a lot of uncertainties in things that we can't control. I think we've got much more control over NIM and costs, which is why we're content to guide on those. We'll do our best in other respects.
Okay. I appreciate there is a lot of uncertainty, so thank you very much for that.
Thank you.
We have a question from John Cronin from Goodbody. Your line is now open.
Hi, Ian. Thanks for the call. Just a few from me. One is on, given the experience, given the balance transfer card market is seeing shrinkage in terms of product availability, how does that now play into your thoughts around retention of balances, and could there be scope for an adjustment in the EIR on those balances, which could therefore boost capital in the short term? The second question is just on pensions. Look, clearly you've agreed a roadmap with the trustees of the pension scheme in terms of contributions for another number of years, but what are the risks on the horizon in that respect in terms of updates and updated actuary valuation and what that might mean? Thirdly, just to follow up on a previous question on deposits current term deposit capture.
Look, any guidance you could give us in terms of evolution of the composition of the deposit book would be helpful, but noting you're wondering whether it could accelerate a little bit faster from here, observing your move last week to bump up the rate for current term customers, albeit just with respect to small balances, but keen to understand how we might think about that given rate attrition in the market more broadly. Thank you.
Hi, John. Thanks. I was really hoping you'd have one for David there, but I'm happy to have a go. Look, I'm sure David will compliment it if I don't do a good job. Balance transfer. Yeah. It's interesting. Again, it's so hard to make predictions at the moment about what might happen. I'm responding to your question, trying to be helpful. Typically, in our sort of stress modeling, normal customer behavior, and this is what we've observed previously, is that in a practice when there are demands on customers' funds elsewhere, and less availability of credit, you do see people stick around post promo. Essentially because there are no better offers out there. That is helpful from a sort of income and capital perspective. It's certainly some of the behavior that we model into our ICAAP stresses and other things like that.
I think it is a reasonable assumption, but it isn't something that we're sort of baking into either margin guidance or volumes or anything else at the moment. I think we're going to have to just see how this plays out. Yeah, it is a sort of basic tenet that when there are fewer offers around, you tend to see, and this is what we've seen in our sort of historical modeling, you tend to see people stick around for a bit longer, and that is helpful from an income perspective, when people are post promo. On pensions. The triennial is going on at the moment. It's actually quite a difficult environment in which to do a triennial valuation given the sort of disruption that's been experienced.
I guess broadly speaking, the funding position has improved every year, and I think we're in good shape from a trustee funding perspective. The conversations with the trustees are really around what does the journey plan look like? Because we're much closer to the end game on this than other schemes given the age and stage of most of our pensioners. Those conversations are still ongoing. The key things or sort of risks or additional funding requirements that would drive the short-term sort of trustee deficit, they're pretty small actually. We're talking about equalization, those sorts of things. Nothing that we think is going to move the needle adversely in terms of capital requirements coming out of triennial discussions. I hope that helps. On deposit capture. Again, it is hard to tell.
We've seen a big uptick in balances on current accounts and other things at the moment. That should iron out. You would imagine that as customers face into one or two challenges, they might spend some of the cash that they've hoarded over the last couple of months. I think for us, the tweak we did last week on the current account proposition was just really about enhancing our overall offer, and it was good to see some of the personal finance editors pick up on what I think is a more customer-friendly proposition all round in terms of both what we'll offer on small positive balances and, I think a superior overdraft proposition for most customers. That might help us. We're not going to be in the market heavy marketing and things at the moment, simply because we're not sure we'll get best bang for buck.
I think the real focus is how do we accelerate current account acquisition once we get into later in the year, both on the business and personal side. That really is about dusting off the original plans and propositions and really getting on and executing them. I think maybe the one change is we've probably learned a lot more over the last eight weeks on how to sell products digitally than we knew before. I think that's a good thing.
It's very helpful. Thanks.
Okay. I think we've exhausted all the questions. We'll kind of close it there. The investor relations teams are around if anyone's got follow-ups. Feel free to get in touch. Operator, I think we're done.
Ladies and gentlemen, this concludes today's call. Thank you for joining. You may now disconnect.