Business, Energy and Industrial Strategy Committee
Oral evidence: Covid loan fraud, HC 1190
Tuesday 15 March 2022
Ordered by the House of Commons to be published on 15 March 2022.
Business, Energy and Industrial Strategy Committee Members present: Darren Jones (Chair); Tonia Antoniazzi; Alan Brown; Richard Fuller; Ms Nusrat Ghani; Mark Jenkinson; Andy McDonald; Charlotte Nichols; Mark Pawsey; Alexander Stafford.
Treasury Committee Member present: Siobhain McDonagh.
Questions 113 - 196
Witnesses
III: Hannah Bernard, Head of Business Banking, Barclays UK; Karl Reid, Head of Government Lending within Commercial Banking, HSBC; Susan Davies, Head of Business Banking, Santander UK; Jas Singh, Consumer and Business Banking Director, Lloyds/Bank of Scotland; Andrew Harrison, Head of Business Banking, NatWest.
Witnesses: Hannah Bernard, Karl Reid, Susan Davies, Jas Singh and Andrew Harrison.
Q113 Chair: We are now going to move on to our third and final panel. We welcome Hannah Bernard from Barclays bank, Karl Reid from HSBC, Jas Singh from Lloyds/Bank of Scotland, Andrew Harrison from NatWest/RBS and Susan Davies from Santander. Good morning and thank you to all of you for appearing this morning. I am very conscious that there are five of you, there are a number of questions from colleagues and we are running against the clock. Could I encourage you please to give direct answers to the questions that you are asked and not read out pre-scripted comments that you might want to deliver? That would be very helpful to us.
First, I am just going to ask a couple of direct questions to each of you. Again, please give me direct answers. If you do not know, please say that you do not know and you will write to me afterwards. Hannah Bernard, first of all, how much did your bank lend in total off the back of Government guarantees through the three main schemes?
Hannah Bernard: If I take bounce back loans first, that was £10.8 billion; the other two I will have to give to you later.
Q114 Chair: How much of that Government-backed funding was used to refinance existing credit facilities at your bank?
Hannah Bernard: We believe that 1% of term lending would have been refinanced on bounce back loans. There will be an amount of overdraft that will have been repaid as well.
Q115 Chair: How much of the money went to companies that had been dissolved?
Hannah Bernard: We will have a very small proportion that will have subsequently filed a notice of dissolution at Companies House.
Q116 Chair: You had no companies that were already dissolved but were able to get access to financing.
Hannah Bernard: We put in place lots of fraud measures. The latest information from BBB has suggested that there were small amounts that were not in our marker set that we got through, and they may have subsequently put through a notice of dissolution.
Q117 Chair: Did you have any duplicate applications?
Hannah Bernard: We did at first, before we got the industry scheme up and running. We did not have any that were duplicate within Barclays, but within the industry we had about 1,500 that were duplicates.
Q118 Chair: Lord Agnew suggested that some banks had dropped their fraud markers, which apparently can be assessed in real time against company applications. Did your bank do that?
Hannah Bernard: We put in place a completely different system. If you can bear with me for two minutes, there were two things we did that were quite important. First, we put the application form behind our firewall. People had to be registered for online banking before they could access the application form. Secondly, we put lots of fraud markers on because we could not use our normal system. We had a different system, a different process. We put markers on the customers. Then 60% of applications went straight through the online process and 40% went via a colleague. We were very vigilant to try to make sure that, where we needed to, we had an extra set of eyes on it.
Q119 Chair: Where you have accounts where it looks as though fraud has taken place, will you be recovering the frauds using the Treasury guarantee or recovering the funds internally?
Hannah Bernard: We will be absolutely following the process that we have agreed to try to get that money back, working with law enforcement agencies and the industry bodies we have. If it subsequently transpires that we cannot get our money back, we will make a claim on the guarantee.
Q120 Chair: Everyone else now knows the questions I am about to ask, so they have an increasing amount of time to get the answers. Jas Singh, how much did your bank lend under the three schemes?
Jas Singh: On bounce back loans, it is approximately £9.7 billion. It is a smaller number on the CBILS scheme.
Q121 Chair: How much was used for refinancing credit facilities at your bank?
Jas Singh: That was incredibly small. The vast majority of businesses that took out bounce back loans had no previous history of borrowing with us.
Q122 Chair: They had no previous history of borrowing at all.
Jas Singh: Yes, these were businesses that, in their typical course of running a business, may not have felt the need to borrow to grow the business. From what we saw, they were predominantly borrowing to pay for short-term wages or as a safety net.
Q123 Chair: Did you end up lending any money to companies that had been dissolved?
Jas Singh: No, not to businesses that were dissolved at that point. In fact, one of the things we put as a check in our own journey was whether any customer or business had been in financial difficulty pre the pandemic. Then we would make sure to stop their application before the bounce back loan was processed.
Q124 Chair: How many duplicate applications were there?
Jas Singh: There was an incredibly small number, actually. There are two things. First, in the way we set up our own journey we check for any duplicate applications. Businesses could only go through the journey once. The second time they applied, it would stop right at the outset. Secondly, an industry-wide database was set up that tracked anyone who had a bounce back loan with any of the other banks. We have tracked that, and those numbers are very small.
Q125 Chair: Did you use your existing fraud markers or did you set up a bespoke scheme?
Jas Singh: Across the board, we used our existing fraud markers. Our fraud markers checked the business that is trading with us. One of the choices we made was that we only made bounce back loans to businesses that already had a relationship with us. We already had checks on their history. We could check, where relevant, whether they were trading on Companies House, and so on. We could continue to use our standard fraud markers.
Q126 Chair: Presumably, if you cannot recover funds from normal fraud schemes, you will rely on the Government guarantee.
Jas Singh: We have done a few different things. First, wherever we have identified fraud, we have gone ahead and recovered the funds from those businesses. In a reasonable proportion of cases, we have been able to recover funds. Secondly, yes, we are working with the Cabinet Office and with NATIS and so on to pursue those businesses. Where we do end up in places where we do not think we can recover the funds, we will rely on the guarantee.
Q127 Chair: Karl Reid, what was the total amount lent under the schemes?
Karl Reid: On bounce back loans it was £7.2 billion. I would have to come back to you on the other schemes.
Q128 Chair: How much was used for refinancing at your bank?
Karl Reid: On bounce back loans it was very, very small. As my colleague said, at HSBC, it was pretty much the case that the companies that were borrowing had never borrowed before. This was the first time they were entering into an agreement. It is quite a small number.
Q129 Chair: Just to be clear, for witnesses who say “very small”, we will expect actual numbers. We can do that in writing afterwards. How many companies that had been dissolved did you lend money to?
Karl Reid: Again, it was a very small number. Clearly, we are working through the various activities to identify those.
Q130 Chair: What about duplicate applications?
Karl Reid: We did receive a number of duplicate applications throughout the scheme. We only lent to a single application in HSBC.
Q131 Chair: If you do not know the number, just say, “I do not know.” I need to move through quickly. On fraud markers, did you have a bespoke scheme or did you use your existing fraud markers?
Karl Reid: We created a bespoke scheme. It was completely different to our BAU lending. We brought in some of the activities that we would have done in our BAU lending; before we disbursed funds, we did additional checks.
Q132 Chair: Presumably you will rely on the Treasury guarantee, if you cannot recover fraudulent applications in the normal way.
Karl Reid: After our collections and recoveries processes, yes.
Q133 Chair: Susan Davies, how much was lent from your bank, please?
Susan Davies: There was £3.9 billion under BBLS to 252,000 companies. There was £700 million under CBILS to about 2,000 companies.
Q134 Chair: How much was used for refinancing?
Susan Davies: It was a very small proportion. I do not have a number. Similar to my colleagues, over 71% of the applicants for BBLS had never borrowed before.
Q135 Chair: How many dissolved companies did you lend to?
Susan Davies: Again, I do not have the exact number, but it was a small proportion. We made the decision to make BBLS available only to our existing customers. We had a degree of visibility with the customers. If we saw anything that was untoward, we could put a stop on before paying out the funds.
Q136 Chair: How many duplicate applications were there?
Susan Davies: Similarly, I do not have the exact number, but again we had some visibility, because they were existing customers.
Q137 Chair: Did you use a bespoke fraud marker scheme or an existing one?
Susan Davies: We used existing fraud markers, because we were dealing with our existing customers. We maintained those fraud markers throughout the scheme.
Q138 Chair: Presumably you will be relying on the Treasury guarantee if you cannot recover fraudulent funds.
Susan Davies: Yes, we stopped £85 million of loans being paid initially through the checks we did. Since then, we have recovered around £20 million through interventions and actions from the data and insight we have had.
Q139 Chair: Andrew Harrison, what was the total amount lent?
Andrew Harrison: Across all three, it was about £14 billion. Around £9 billion of that was bounce back loans.
Q140 Chair: How much was used for refinancing?
Andrew Harrison: That is not a figure that we have tracked. Similar to everybody else, 70% were not borrowers pre obtaining the loan.
Q141 Chair: Did you lend any funds to companies that had been dissolved?
Andrew Harrison: From the information we have so far, we believe there are about 100 firms.
Q142 Chair: What about duplicate applications?
Andrew Harrison: It is a similar amount. There are none within the bank, but with other lenders. About 100 is our current view.
Q143 Chair: Did you use existing or bespoke fraud markers?
Andrew Harrison: We used our existing internal fraud markers. We do Cifas checks, which is external, plus we were also running analytics through a couple of systems to check for fraud behaviours before any loan was drawn down.
Q144 Chair: Presumably you will be relying on the Treasury guarantee for funds that cannot be recovered.
Andrew Harrison: If we follow the scheme rules, that will be the case, yes.
Q145 Chair: I know that all of you were asked to do a great deal of work quickly and you had very significant applications. Ultimately, this was quite a good deal, was it not? You can recover funding from Government if you need to and you are able to refinance existing credit facilities. None of you gave me an answer specifically about how much was used for refinancing in your own bank. Is there a reason why you do not have that number today? Is it because you are not collecting the data or because you do not have it to hand? Are you collecting the data? I can go back through again—yes or no.
Andrew Harrison: It was within the scheme rules. The majority of borrowers that came through were non-borrowers previously. It is quite hard to track what has gone to a deposit account and what has gone to repay an overdraft, because overdrafts might be drawn and repaid on a regular basis. It is just not a figure that we have been monitoring.
Chair: Are there any other contributions on that?
Jas Singh: The way the bounce back loans worked is that someone will have got money into their business current account. It would have been very sensible in some cases for those businesses to use that money to pay off their existing overdraft, because they were trying to trade through a pretty critical period. That would have made complete sense. In fact, I would have thought that it would not have been great from a scheme design perspective or from the banks’ perspective to say, “You have cannot use this money to pay off overdrafts.” Similarly, the same thing applies to any loans they might have. They might have payments coming up for the loans in the next few months. If they use that money to pay that off, that should be completely fine from a scheme design and a business outcome perspective.
While there are different interest rates that consumers or businesses pay for the loans they have, in a reasonable majority of cases they will have been paying higher interest rates on other products. In some sense, if you think about it from a bank’s incentive perspective, someone who is paying a higher interest rate ends up paying a lower interest rate. There would not be any active intervention from us to say, “Yes, please go ahead and do that.”
Q146 Charlotte Nichols: As experienced lenders, you would have had established means in place to give out loans to businesses. From what we have heard so far, many of you were lending only to existing businesses. Given this, how did you advise or make suggestions to the Government on the way that the covid business loans were set up? Could things have been done better?
Andrew Harrison: Very constructive discussions were had with the British Business Bank, Treasury and UK Finance, which was representing the lenders through that period of time. There were open discussions around the balance of risks and the desire to get money to businesses very quickly. At that time, many reports were saying that businesses had as little as two weeks of cash flow left, so there was a desire to do that. As we have heard in previous evidence, there was a good understanding of some of the risks that were involved in terms of doing that.
From our position as an institution, we very much made sure we put in place the fraud protocols up front to protect the taxpayer as much as possible. Indeed, in many ways we went beyond what we would normally do with our traditional lending by putting many of those checks up front. We screened hundreds of thousands of applications that did not get through the front door, because our checks stopped them. We put appropriate checks in place to protect the taxpayer through that very intense period of time.
Susan Davies: As lenders, we recognised very early on that the CBILS scheme was not going to be suitable for many of the small businesses that really needed that lifeline. As many people know, it was based on the old enterprise finance guarantee. Understandably, it was a suitable mechanism for larger businesses, but you had to do the due diligence. For many businesses that had never borrowed before, that just was not going to be a feasible option. As lenders, we were very vocal about that. We worked collaboratively with the industry to find what was going to be a better and more suitable mechanism. Of course, that was BBLS.
Karl Reid: If we think back to what was happening, the economy was shutting down. We were working very closely with the Treasury and British Business Bank in designing the scheme. We had a very short time to implement that. We need to be aware of the very tight timescales we had. We did the necessary fraud checks. We also ensured that our internal fraud protocols were completely reviewed and understood. We did stay open to new-to-bank, but for every onboarding we did we ensured that we followed our business-as-usual onboarding checks, which included know your customer, anti-money laundering and credit checks, before we progressed on to a bounce back loan.
Jas Singh: I will not repeat the points about speed and urgency, because my colleagues have made those. Having been in some of those conversations right at the outset, the fact that we were talking about lending taxpayers’ money was not lost on the participants. There was active consideration by us, other banks, the British Business Bank and HMT to make sure we were doing everything we could to balance the different competing priorities: how do we make sure businesses get money in the short term and taxpayers are protected?
While a lot of the thinking and execution was done at speed, there was a lot of good engagement. The scheme was launched on 4 May, and a couple of days later we got together on a call. There was UK Finance, the British Business Bank and all the banks that were participating in the scheme. Two days later, we had started to share what each one of us was seeing in terms of concerns around controls. Having been in those conversations, I feel we were quite thoughtful in thinking about the implications, and acting and intervening quickly.
Hannah Bernard: The speed and scale thing should not be lost. If I can just give the Committee a fact, within the first four weeks we did six years’ worth of volume of lending. I hope that puts this into context. Yes, we would have all loved to have had another four weeks, say, of preparation, but we would then have been slow in getting the money to those hundreds of thousands of businesses that desperately needed it. When you talk to businesses now about their pandemic experiences, the bounce back loans were a lifeline for vast swathes of businesses.
Q147 Charlotte Nichols: No person on the Committee is disputing that. What we are trying to get to is what discussions were had with the Government, how willing they were to take on board some of your suggestions and, if there were suggestions they did not take on board, whether that is something that has caused some of the fraud and loss we are seeing. I do not mean to cut across you there, but I am interested in hearing whether there were things that were suggested that were not taken up or whether there were things that were said that were not done. Were the Government, as some of the other people on the panel have said, taking those things on board and having quite an active dynamic relationship with you?
Hannah Bernard: If I could give you one example, the scheme necessitated a lot of self-attestation, particularly on the business’s turnover. Ideally, we would have wanted an HMRC data feed so we could have checked the turnover they were stating. That was not available. We made that suggestion; it was not available. We had to go with self-attestation.
Q148 Chair: Why was it not available? Did they not provide it already? Was it a bespoke new ask?
Hannah Bernard: Yes, I do not believe it was available.
Q149 Mark Pawsey: I am struggling a bit here. We heard from Lord Agnew about substantial levels of fraud. We had an acknowledgment from the British Business Bank that fraud has occurred. We have the representatives of five major banks telling us that they employed all their usual checks and there were not major levels of fraud. Who is right and who is wrong?
Hannah Bernard: It is important to think about the different categories of fraud. The first category is organised crime. The second category is scheme abuse. You have heard about the multiple bounce back loans. That is scheme abuse. The third category is around self-attestation. When you hear these big numbers, it is the self-attestation piece that falls into that category of fraud. That does not necessarily mean those people are not going to pay us back. That is important, because traditional fraud is still relatively low. It is the self-attestation piece that is the thing that is potentially much larger.
Q150 Mark Pawsey: As bankers, do you contest the levels of fraud that we have heard about in our evidence session previously from Lord Agnew and acknowledged in part by the British Business Bank?
Hannah Bernard: It is still really early days. We have seen relatively low levels of traditional fraud.
Q151 Mark Pawsey: I am struggling with “traditional fraud” and “other fraud”. Fraud is fraud, is it not? This is people stealing money from the taxpayer.
Hannah Bernard: If I just take the self-attestation piece, people will have made estimates about what they thought their turnover was going to be. That does not necessarily mean they were intending to defraud and not pay us back. If they have deliberately inflated their turnover, I completely accept that would be fraud, but they may still pay us back.
Q152 Mark Pawsey: Under your usual checks, when a business stated its turnover, you would check that—
Hannah Bernard: We would do.
Mark Pawsey: It is available from Companies House. How can you give all these loans to people who were making up figures about their turnover?
Hannah Bernard: We weren’t to do credit checks, because we were trying to get the money—
Q153 Mark Pawsey: I do not want to put you on the spot, Susan Davies, but we heard from you that you put your usual checks in place.
Susan Davies: We did, yes. Just building on my colleague’s point, in our case in Santander, we did see a relatively small amount of fraud—less than 1%. We have continued to see those levels throughout. The experience in terms of the types of fraud is similar.
Q154 Mark Pawsey: You did not have to put all your usual checks in place. That was the purpose of the Government scheme: to get the money out of the door quickly. I am unsure why, as bankers, you are telling me, “We used our usual fraud checks to make certain these things did not go wrong.”
Andrew Harrison: The design of the scheme was such that, to accelerate the funding going to businesses, the businesses would self-attest on things like turnover. The second element is that the banks were instructed not to undertake credit checks to verify things like turnover and so on.
Q155 Mark Pawsey: That is anathema to your normal way of doing things, is it not? Did you do them anyway?
Andrew Harrison: That would delay the proceeds getting to businesses. The banks were instructed to undertake their normal fraud checks.
Q156 Mark Pawsey: You have told us that in most cases you were lending to people who were already your customers. You knew their history.
Andrew Harrison: We knew their history, and we undertook the right level of checks to understand whether those individuals who were borrowing had behaviours that indicated there was fraudulent activity in place. We were instructed not to undertake the normal credit checks, because that would have precluded a large number of businesses getting access to those funds.
Q157 Mark Pawsey: How did you come to understand the operation of the rules? Did you get emails on a regular basis from the British Business Bank? How did you know what the terms of these different deals were?
Karl Reid: We have the lender’s manual.
Q158 Mark Pawsey: Who prepares that?
Karl Reid: That is from the British Business Bank. It outlines the guiding principles of the scheme. We also have the guarantee that lists out the criteria.
Q159 Mark Pawsey: In terms of the route of information dissemination, it may have come from the Treasury or a Government Department, but it then comes to the British Business Bank. As the lenders who are putting the money out to these businesses, your relationship was always with the British Business Bank.
Karl Reid: That is correct, and obviously with our customers, because we are always servicing our customers.
Q160 Mark Pawsey: Yes, but you were not able to influence Treasury; you did not have any discussions with the Cabinet Office or BEIS. What happened there was irrelevant to you as banks, because you were told what to do by the British Business Bank.
Karl Reid: We were in discussions, as part of the industry forum, as my colleagues have said, with UK Finance. We were engaged with the British Business Bank and HMT in the early days of the scheme, before the scheme was initiated.
Q161 Mark Pawsey: Did you have any opportunity to influence the schemes on the basis of your experience of lending to businesses?
Karl Reid: We provided our input into the various discussions, but, ultimately, because of what was happening in the economy at the time—
Mark Pawsey: It was the speed.
Karl Reid: It was the speed. The other big differential, which my colleagues have not discussed, is affordability. One of the big checks that we would do in our normal business-as-usual lending is an affordability check. We were not allowed to do that affordability check as part of the bounce back loan scheme.
Q162 Mark Pawsey: Say that again. You were not allowed to do that check. It was not that you were discouraged from doing it; you were not allowed. Who would not allow you to do it? Was it the British Business Bank?
Karl Reid: For the CBILS scheme, as we have spoken about before, we did all the necessary checks that would follow business as usual, which included affordability checks. That would add several weeks on to a loan application.
Q163 Mark Pawsey: How would it add several weeks to that, if the clients were known to you?
Karl Reid: Because you still need to look at their business plan. Part of assessing any credit is assessing the customer’s ability to pay in the future. It is not just looking backwards but also looking forwards.
Q164 Mark Pawsey: Does that have to take weeks?
Karl Reid: We work with our customers to assess each credit application, as we did with CBILS. In bounce back loans, as we were saying, it was imperative to get cash out to customers at speed. Just to give you an example of a customer, we had an independent clothing chain that was forced to close. It had seven employees. We had to ensure it could get the money to pay the employees, which was very important. One of the other suppliers it had to pay was HMRC. Helping customers like that to continue to be able to transact and pay their employees was very important.
Q165 Mark Pawsey: I might ask Mr Singh this. Lord Agnew placed great store on the fact that some of these loans were being used to pay off other loans, be it credit card debt or whatever. You do not see any problem with that.
Jas Singh: As I said earlier on, the scheme was designed to make sure we were helping businesses trade through a very tough period. If what those business owners needed to do was pay off an overdraft or pay wages, I would have thought that was okay from a scheme perspective.
Q166 Mark Pawsey: Why did Lord Agnew think that was improper use of these loans?
Jas Singh: I am not sure. I cannot speak to that.
Q167 Alan Brown: Mr Reid, Lord Agnew told the Treasury Committee on 9 March that HSBC had taken four months to put duplicate checks in place for businesses applying to the bounce back loan scheme. Is that four-month estimate correct? If so, what were the reasons for the delay?
Karl Reid: We need to go back again to what was happening at the time. At our peak, we had in excess of 30,000 applications a day coming into HSBC. We were there to try to service our customers as quickly as possible. Therefore, we were doing the necessary checks. For onboarding of new customers, we brought the check in as soon as possible.
Q168 Alan Brown: How quick is “as soon as possible”?
Karl Reid: For new customers, we brought it in at the end of June. Then there was a delay in ensuring that we could roll it out to the wider population. At the time, there was a lot of pressure around ensuring that the loans were being disbursed in 24 hours.
Q169 Alan Brown: You said there was a delay. Did it take those full four months to implement the duplicate checks?
Karl Reid: We eventually went live with the full duplicate check on 8 December. However, we were providing data to the Cifas FIND check as of May. We were uploading our data to that facility.
Q170 Alan Brown: How long did it take each of the other banks to get their duplicate checks up and running?
Hannah Bernard: We were running by mid-July.
Alan Brown: How long was that?
Hannah Bernard: It was six weeks from the start of BBLS.
Jas Singh: It was very similar for us. The duplicate information was available from around the last week of June. We were ready by the first week of August with our checks.
Susan Davies: Similarly, we were early July. It was set up at the end of June.
Andrew Harrison: Pretty much as soon as it came out, we implemented it. It was a couple of weeks.
Q171 Alan Brown: I will go back to you, Mr Reid. Why was HSBC so far behind the other banks in completing that?
Karl Reid: Like I said, we were trying to get the funds out to our customers as quickly as possible. We also maintained an open status for new customers. I am proud to say we were able to open 71,000 new accounts, and we also—
Q172 Alan Brown: Sorry, does that mean that other banks were at risk of exposure because HSBC was not doing the duplicate checks quickly enough?
Karl Reid: As I said, we uploaded all of our information on all our applications to Cifas FIND from the end of May. The other institutions will have been able to see our information.
Q173 Alan Brown: On the actual process of doing the checks yourselves, it still took longer for you to implement these checks.
Karl Reid: We implemented the checks as soon as operationally practical for us.
Q174 Alan Brown: I am confused about this. You mentioned 8 December, doing it right away and the information being available. Either duplicate checks were getting done right away or they were not. Was there a delay in the duplicate checks being undertaken for a certain cohort of loans that you were giving out?
Karl Reid: As I previously said, for new-to-bank customers we started those checks at the end of June. We put the wider check in as soon as was operationally practical for us.
Q175 Alan Brown: When was that?
Karl Reid: It was 8 December.
Q176 Alan Brown: That still seems a while to me. If we get sight of the fraud dashboard, is it likely to show that HSBC does have a bigger exposure—or, ultimately, that the taxpayer has a bigger exposure, through HSBC—as a consequence?
Karl Reid: I do not have information on my other colleagues and what they are doing. However, we are working through that population. We have seen that customers have actually repaid these facilities. We are working with each of the customers, where we see a duplicate loan.
Q177 Alan Brown: Earlier on in evidence we heard the representatives of the British Business Bank saying that they have revised the estimates of fraud levels from 11% down to 7.5%. We can look at that as either good news or still really bad news. Could each of you give me a figure for the fraud level that your bank has been exposed to?
Hannah Bernard: In terms of the—
Alan Brown: It is just the figure.
Chair: In the interests of time, we just need a number or “don’t know”, if that is all right.
Hannah Bernard: The reason why it is important is that it is as at today, what we have in the portal, versus what we think relates to the 7.5%. As at today, we have 1.5%. That includes traditional fraud and the self-attestation piece.
Jas Singh: We do not recognise the 7.5% number. We have a much smaller number than that. It would probably be in the 2% to 2.5% range.
Karl Reid: Currently, it is 1.81%.
Susan Davies: It is less than 1%.
Andrew Harrison: We think it is about 2%, given the information we currently have.
Chair: We might come back to that.
Q178 Alexander Stafford: Given that the liability for the losses sits within the Government, how much effort have you guys made to recover any fraudulent money? Surely you can sit back and say, “It is nothing to do with us.” You might keep saying that these are small amounts—1%, 2% or 3%. That is still a lot of money. Surely you have the responsibility to try to get that back. What have you done? Have you gone back through the applications? How much work has each of your banks put in to try to recover this money rather than just passing it off to the Government? How many people are working on the team looking at this, for instance?
Susan Davies: An enormous effort continues to be focused on recovering what is ultimately taxpayers’ money. We have worked collaboratively across the industry. We get information from BEIS in terms of being able to identify potentially fraudulent customers. I know at first hand that, on a daily basis, my teams are reviewing these cases, going out and contacting customers, and following through these cases to conclusion. While we recognise that no instance of fraud is acceptable, we have been able to recover nearly £20 million through the efforts of the teams. They continue to focus on that activity to this day and they will continue until we have the opportunity to recover as much money as we can.
Q179 Alexander Stafford: You mention an “enormous effort”. That means nothing to me. I do not know what that means. Everyone is making an enormous effort. Have you gone back through each of the applications from businesses to find out where the fraud is? How many people are working on it? While £20 million sounds like a lot of money, there is a lot of money—billions—being sent out. Out of however many billion you have lent, £20 million is not 1%; it is smaller than that. How much effort have you and all your teams been making to try to get this money back for the taxpayer?
Susan Davies: As I said, we are making a concerted effort to—
Q180 Alexander Stafford: How big is your team?
Susan Davies: We have a team of over 100 people who are dedicated and focused on fraud activity. There is a much larger team behind that. That is a team that will focus primarily on the data and the information that is provided to us.
Q181 Alexander Stafford: How many people are working on it? How many did you have working on fraud prior to covid?
Susan Davies: Unfortunately I do not have that number.
Q182 Alexander Stafford: Has the team increased? If you are giving out far more money, surely the fraud team should increase to account for it and to find the fraud. Has it remained the same? I would ask that everyone answer the question about how much your fraud teams have increased. Have they remained the same, grown or shrunk? Do you have the numbers?
Andrew Harrison: In terms of the energy and effort, first of all, the vast majority of the fraud or potential fraud was stopped at the front door before it got through, because of the checks. As I say, our currently observed fraud is around 2%. We have recovered through activity probably around a third of that. It is important to say that, of that 2%, we also have customers within that cohort who are still repaying. It is suspected fraud, but it is not necessarily manifesting itself in losses.
In terms of what we have done, similar to what Susan was saying, we have put in place increased resources. We probably have about 100 additional people within our teams focusing on that. We have also invested in additional analytical tools that we have run right across our back book to make sure we are looking for triggers and behaviours that might indicate fraudulent activity. Where we do that, we follow that up with the customer. We are also working very closely with law enforcement agencies, the industry, the Cabinet Office and others to follow up on suspicious activity that we have identified. Off the back of that, we exited about 260 businesses that were involved in organised crime, which was part of a wider exercise across the industry. It is a very significant activity that is under way.
Q183 Alexander Stafford: You have not answered the question. Has your fraud team increased?
Andrew Harrison: Yes, by around 100 people.
Hannah Bernard: We have increased our fraud operations team by 20%. To Andy’s point about the up-front controls, at its peak we had an extra 900 people working on processing fraud markers.
Q184 Alexander Stafford: You said your fraud team increased by 20%, yet earlier you mentioned however many more loans there were. It was six years’ worth. Surely it should be six years’ worth of fraud people investigating, if it were to scale.
Hannah Bernard: No, to Andrew’s point, a lot of the volume was electronic and physical applications up front. In terms of the number of cases we are now doing at the back end, that is where we have increased our fraud team.
Jas Singh: At its peak, we had about 250 people working on fraud detection. There would have been about 100 doing similar activity in the past before the pandemic. The other thing to bear in mind is that a lot of fraud detection does not happen just with people; it happens with technology. Across all of us, there has been lots of investment in the journey up front but also the ongoing monitoring.
Karl Reid: We have ongoing account and transaction monitoring. If there are any triggers on the back of that or if any of our staff see any unusual activity, we would raise an unusual activity report, which would then be sent to the fraud operations team. If there were any suspected serious fraud, we would raise a suspicious activity report, which would go out to NATIS and other enforcement agencies. We have increased our fraud team. It is dynamic, and it is something I continuously look at. If we need more, I will bring more people into my team to make sure that is covered.
Q185 Richard Fuller: In the interests of time, I will ask specific questions of specific people, if I may. Mr Harrison, how many bounce back loans did NatWest give out in total?
Andrew Harrison: It is around 300,000.
Q186 Richard Fuller: In response to the Chair’s question about the number of duplicates, you said there were about 100 companies. I was interested in that. It is always interesting to look at numbers and the difference in how you present things. Out of 300,000, 100 is about 0.04%. That is not a very large percentage. Similarly, from the information from the British Business Bank, there were something like 1,500 companies that were dissolved, duplicates or incorporated post. In total, that is about 0.175%. Is that correct?
Andrew Harrison: Based upon the information we have today, yes.
Q187 Richard Fuller: It is interesting that the percentage, in terms of the number of applications, is small. Would you say that is a relatively small percentage, just a small percentage or a big percentage?
Andrew Harrison: It is hard to tell, because there is no precedent to look behind. Any loan that goes out incorrectly is something that is unsatisfactory. It highlights that there was a good degree of checks in place to stop that.
Q188 Richard Fuller: Just to be clear, between 24 May and the end of June, so one month, banks received a million applications. I think that statistic is correct.
Andrew Harrison: Yes, I do not have that exact detail.
Q189 Richard Fuller: Ms Bernard, can I just talk to you a little bit about the interesting point on the percentage of fraud? As my colleague Mr Brown was saying earlier, the estimate has come down from 11% to 7%. We went through each of the banks and you were at about 1.5%. What accounts for that substantial difference between how the banks are estimating fraud and what is in the domain from Government estimates?
Hannah Bernard: It is timing, to be honest. This is our point in time today. We are still relatively early in the journey. When I look at the third bucket, which is about the self-attestation piece, that may still manifest itself when people fail to make payments. When we go back and check, it may look like they have inflated their turnover. If we do an analytical review of what that might be, I can see my way through to the 7.5%, but most of that—90% of that—will be self-attestation, not duplicates or the scheme abuse piece we have been focusing on today.
Q190 Richard Fuller: Rather than us hoping that you will be bearing down on that 1.5% to get it to 1% or 0.5%, your anticipation is that the 1.5%, over the months to come, will go back closer to the 7.5%.
Hannah Bernard: The 7.5% will be if people are not paying us back. If everyone who has inflated their turnover did not pay us back, it might go up to that. When I look at the book today, there are people who we think may have inflated turnover who are paying us back.
Q191 Richard Fuller: They have cheated and got away with it.
Hannah Bernard: There are a couple of things, actually. If I am a sole lender and someone has two bank accounts, one with one of my colleagues here, I can see only the part of the turnover that they have with me. It is possible that we are all looking at the book and going, “That looks like it is a little bit high,” but actually we do not have the full picture. That is why the HMRC feed would have been useful. They are the only people who will know what the turnover is. If they continue to pay—if they have taken a loan and they pay it back—it will not get categorised as fraud.
Richard Fuller: There is an interesting minefield in terms of what people define as fraud in those circumstances, I guess.
Hannah Bernard: It is an interesting moral question.
Q192 Richard Fuller: Yes, very. Mr Singh, if I could pick on you for this question, a number of points were made earlier to us about public disclosure of the companies that have received bounce back loans and that there is really no issue about that being made publicly available. Do you agree that there would be no concern if the Government were to ask you to make that information available on behalf of the companies to which you have made bounce back loans?
Jas Singh: Was the question about whether we would have any concerns disclosing the different levels of fraud across the different banks?
Richard Fuller: No, across all recipients. For every single company out of the hundreds of thousands that you may have given bounce back loans to, should that information be in the public domain?
Jas Singh: I am not sure what purpose it would serve, but, if it were the decision of the British Business Bank and the Treasury that it would be helpful to disclose everyone who received a bounce back loan publicly, we would happily support that.
Q193 Richard Fuller: I want to be clear. On behalf of your customers, you are saying that, if that disclosure were required, you would not be making representations to say, “Hold on a minute. There will be a problem if we have to do that.”
Jas Singh: Again, I am not sure what purpose that would serve.
Q194 Richard Fuller: The purpose would be what the Government see. Let me give you one example of purpose. We have asked taxpayers to put out £47 billion or more and some of that may have gone missing. The public has a right to know. That is a pretty good reason to know where the money went. The counterpoint would be if you, as an experienced lender, felt your customers’ businesses’ prospects might be harmed by the very nature of disclosing that they had received a bounce back loan.
Jas Singh: We would have to go back and check with those businesses as to whether we have the right, from a data privacy perspective, to disclose that information publicly. I will need to go away and check whether our current terms with our customers and businesses give us the optionality to disclose their names publicly.
Richard Fuller: I am sorry, Chair. I did say that I was not going to ask everyone. Does any bank here have any particular concern they would like to raise, if we were to make that recommendation that the Government should require that information to be publicly available? No, okay.
Q195 Andy McDonald: I just have a very fast one. In terms of the tools available to you to pursue money where it has been fraudulently obtained, you mentioned the HMRC data feed that was not available. Are there any other asks that you would have of Government in terms of your ability to do that better? Is there anything that is stopping you doing it that you could identify and say, “This is the ask we make of Government. Give us these powers and abilities to pursue it. We could get this fraud debt down further”?
Jas Singh: More rapid availability of information would help. If, for whatever reason, someone who had a bounce back loan was looking to dissolve the company and dissolve the legal entity, us finding out sooner that it is happening would definitely be helpful. The other thing that would be helpful to find out is if we could get access to Companies House at scale. Currently, we have to go through Companies House on a sequential basis. This is similar to Ms Bernard’s previous point. If we had access to Companies House information at scale, that would be helpful.
Hannah Bernard: If there were a centralised enforcement unit, we would get economies of scale through that. That is something we have been talking to the BBB about.
Q196 Chair: Lastly, just to confirm, I am assuming that all your banks are feeding in data to the dashboards the British Business Bank were mentioning. You are all nodding. Is that on a real-time basis? Yes, you are all nodding. When we get to see that, we will be able to see that live information from you.
The only reason I ask is that I am as confused at the end of this session as I was at the beginning. At the start of this session we were told that this is an enormous problem, with up to 25% anticipated fraud and a huge loss of public money. The British Business Bank has said, “It is not that bad, but maybe it is between 7% and 10%.” You guys have come in and said, “Maybe it is about 1% to 2%.” At the end of these two hours, I genuinely have no idea how much fraud there is and how much exposure there is for the Treasury to pay you guys back for these guarantees. We will probably continue to be in touch with you on that basis, if that is okay.
For today, thank you so much for giving evidence to our Committee. We appreciate it.