[2025] UKUT 00185 (TCC)

On 3 August 2018, Mr Stokes emailed Mr Donaldson,saying:

“wehave got to a conclusion on how we should be allocating capital to BTL/PBTL and within both what is the treatment of HMO’s…Impact is that we will have to likely allocate more capital to the back book as the element that needs to move from current 35% RWA to 100% RWA will increase as what has been classed as mortgage lending becomes commercial IPRE [income-producing real estate] lending.”

He went on to say that where a PBTL loan was for over five properties or was a HMO [House in Multiple Occupation], it should be risk-weighted at 100%, but that the remaining loans should remain as residential and thus risk-weighted at 35%. Mr Stokes also noted that the Bank didn’t capture the number of properties on which the loan had been made, or whether it was a HMO, and asked “how do we classify this going forward”, adding that the Bank needed to review “the back book” in relation to classification.

Following a meeting at the Bank on 6 August 2018, Mr MacLean emailed Ms Gillan, Mr Somers and others with the actions which had resulted, which included:

On 7 August 2018, Mr MacLean emailed Ms Gillan, Mr Somers and others, suggesting that the RWAs for commercial lending depended on the loan-to-value ratio (“LTV”) and could thus remain at 50% for some of the loans. Mr Somers replied the same day, saying:

“No, I don’t believe 50% RW is a valid choice as the PRA has used its derogated powers to deviate from standard CRR and adopt 100% for property secured on commercial property.”

On the morning of 17 August 2018, Mr MacLean emailed Mr Arden, saying:

“We have done further work on the RWAs. I think we are much closer to having a minimum estimate for the impact of the change in risk density…I think therefore we need to resolve this issue materially in September.

The key points of progress are:

 We have been through some challenge on the rules and I think we have a joined up understanding. The outcome is at the higher end of the impact ranges we have discussed as pretty much every commercial loan that is not PBTL is 100% RWA classified (or 75% if SME)

 Magdalena and Olivier in the credit risk analytics team have been working to harmonise numbers and I don’t believe we have material disagreements

 We still have an outstanding question re £1.7b of PBTL (whether it is in fact PBTL). The credit team are testing a sample of 170 contracts to test our expected level of accuracy.”

On the afternoon of 17 August 2018, Mr MacLean emailed Mr Arden; the email was headed “data quality issues impacting Reg reporting” and said:

“We have talked in the past about some of the issues that occur in our reg reporting, and how they are impacted by both the historic manual nature of some of our processes, and we have talked about the investments we are proposing that will help with that situation.

We have also flagged data quality as an issue, but haven’t really got into the detail of that in our conversations. At the moment, we are continuing to find issues with our upstream data that have caused inaccuracies in reg and external reporting…We are dependent upon people elsewhere in the bank capturing the right grain of data, capturing it accurately, and not making changes that impact our processes.

The document attached gives a few of the material examples of this problem. I believe that we are in a situation where the risks we are running are now too high. I am keen for us to undertake a piece of work that ensures we have the right processes and controls in place to capture the right data and ensure the data quality is reliable…I think this is an end to end and complex problem and therefore finding a solution might be like eating an elephant. I think the first step is for us to highlight the need and importance of the data that we rely on, and would like your help in ensuring this kind of thing is viewed with sufficient priority.”

The “document attached” included the following, under the heading “misclassification of Loan Facility Type”:

“An extra field ‘Facility Type’ (loan purpose) was added to the CRM data at the back end of last year to close an audit point. An extensive exercise was carried out by the relationship manager all across Metrobank to populate this field in the CRM system to enrich the data with the ‘loan facility type’.

Following on from this exercise, Credit Risk and Regulatory Reporting teams have performed some sampling testing on the data to check the accuracy of the data populated and we have encountered data errors. For examples:

(1)

Credit Risk have picked the top two largest exposures classified as PBTL and checked the underlying collateral secured by these loans. It appeared that these two loans classified as BTL are secured by commercial properties but have been classified as BTL.

(2)

Regulatory Reporting team have also performed some sampling testing and have observed loans secured by residential properties being classified as ‘Commercial Owner Occupier’.

Impact: This observation has resulted in us questioning the percentage of accuracy of the data populated upstream. Regulatory reporting process place a heavy reliance on data accuracy coming from upstream data. We are not able to calculate the risk weights correctly if the information we have place reliance on is not populated correctly.”

Under the heading “Source Data – PBTL Flag does not exist in the system to support regulatory reporting process”, the text read (where “Pepper” is a reference to an external computer system used by the Bank to supplement T24, its main system):

“Following on from the review of the Credit risk weights by KPMG, it has been highlighted to Metrobank that Professional BTL (PBTL) will need to be risk weighted at 100% instead of 35%. However, PBTL flag does not exist in the current Metro system and therefore it is impossible to distinguish between a BTL and a PBTL loans in order to ascertain the PBTL impact on RWA.

Currently, we have about £1b of Pepper and T24 loans that have been flag[ged] as BTL. In order for us to be able to risk weight Portfolio BTL products correctly at 100%, we would need a flag in the system to help us to identify the PBTL loans in the T24 book.

Impact: This will result in the understatement of the RWA numbers until such time Metrobank can start to invest resources and extensive staff time to review the loan documents to ascertain what proportion of the current BTL risk weighted at 35% that will need to be reclassified as PBTL at 100% risk weight.”

On 24 August 2018, Mr Somers emailed Mr Arden and Mr Donaldson, saying:

“Please find attached a short note that details the RWA impacts on two key changes (CRE risk weights change and PBTL definition refinement) required to bring our RWA calculation into compliance. The impact is a circa £900m increase in RWA across Commercial and PBTL books. This represents a circa £70m increase in T1 capital.

Mark Stokes has suggested that a meeting to discuss with you next week might be helpful for context.”

Attached to that email was a PowerPoint presentation headed “RWA Calculating Review–August 2018”. The first slide set out the Executive Summary, which read (our emphasis):

“In September a paper will be brought to CRPAC to advise ELT [Executive Leadership Team] of inconsistencies in current RWA calculations that will result in a significant increase in RWAs.

There are two key drivers of the increase (All RW are quoted before potential SME factors):

1.

Assets backed by commercial real-estate are currently in Metro allocated a standardised risk weight of 50%. This is based on a simplistic interpretation of the European CRR rules. Following detailed PRA statements and reviewing BIPRU confirms that the PRA have used their permitted powers of derogation to ensure that relevant assets in the UK backed by commercial Real Estate should receive a 100% RW. This interpretation has been confirmed by a full KPMG review.

• As a result of this reclassification we estimate that RWAs increase by £640 million (June month end).

2.

PBTL assets are backed by residential properties and currently receive a 35% risk weight. Stratified random sampling however reveals that circa 37% of the balances in the book are actually secured on Multi-Family Dwellings (many leases on a single property), Houses in Multiple Occupation or Student accommodation. These should received a 100% risk weight.

• As a result of this we estimate that RWAs will increase by a further £269 million (June month end).”

Of the £640m figure, £574m related to CLIP loans. The figures had been worked out by the Credit Risk team who had built a calculator to carry out the exercise. The accuracy of the calculator had been verified by first running the figures using the current 50% risk-weighting before running them again using a 100% risk weight for commercial mortgages, and making other more minor changes.

A later slide in the presentation said that “the PBTL portfolio assessment has been on a sample basis as we do not capture planning use class or the number of holders of short hold tenancies as data items” and that the sample was of 100 out of 733 loans, which had been selected on the basis of:

Mr Arden said in his witness statement that this was the first time he was informed that (our emphasis) “the wrong risk weighting may have been applied to CLIP loans and other commercial assets”. However:

We thus find as a fact that before 17 August 2018, Mr Arden was aware that there “was” an issue with the risk-weighting of the CLIP loans, not that there “may” have been an issue.

Mr Donaldson responded to Mr Somers on 24 August 2018, agreeing to his suggestion of a meeting, and adding “we need to consider how we can minimise/offset impact as much as possible”.

The meeting to discuss RWAs which Mr Somers had suggested took place in the morning of 3 September 2018. Mr Arden attended, along with Mr Stokes, Mr Somers, Ms Gillan and Mr MacLean, and a new version of the PowerPoint presentation previously produced by the Credit Risk team was discussed. The two versions were essentially identical, other than that the Executive Summary now included the words: “There is limited potential mitigation in trading book assets with residential property security but this would need case by case review”. In addition, an extra slide headed “decisions required” had been added. Those decisions included the following:

“A decision to recognise the section of corporate loans that will be moving to 100% RWA and to release this in September returns and in the forthcoming ICAAP. We believe the PRA are expecting us to announce this and we risk significant regulatory scrutiny if we fail to act…

A decision on the PBTL book (following from the sampling exercise that defined roughly 37% as arguably not PBTL) to either recognise that portion of the book at 100% RWA or delay (but accept that there is ultimately going to be a requirement under IRB to absorb this capital hit…

In hope of shifting some commercial trading loans currently RWA 100% to 35% there will need to be a sampling of the commercial loans portfolio to determine what proportion may be secured on purely residential properties.”

The contemporaneous minutes record that

In the afternoon of the same day, Mr Richardson emailed Ms Gillan, saying that the Bank had “failed to spot” that the PRA required CLIP loans to be risk-weighted at 100%.