The Bank of England publishes a variety of data that serve as indicators of loan performance. These data show the number or value of loans extended by UK lenders where there are concerns about full repayments being made. These data inform the Bank’s supervisors and policy committees about the health of lenders’ portfolios. They also provide an indication of the level of financial stress in the economy.
This article aims to help users identify the most appropriate loan performance data for their purpose. It explains the different indicators of loan performance published by the Bank. It then discusses other features of the data, including the sectors that the data are broken down into, coverage, timeliness, and data history.
Indicators of distress
The indicators discussed here represent different stages of the loan/asset distress cycle. Lenders make provisions against assets when the transaction first enters their balance sheet, and the amount of provisions may fluctuate over time to reflect changes in the likelihood of repayment. If a borrower misses an obligated repayment they may fall into arrears. If obligated repayments continue to be missed, the collateral the asset is secured against may be seized (possessed) and the asset may ultimately be written off. The different indicators of loan performance published by the Bank, and their sources, are described in more detail below.
Provisions: Provisions represent the funds set aside by lenders to cover potential future losses on outstanding assets. Provisions are made for all assets that are deemed to be at risk of being written-off by a lender. General provisions relate to anticipated future losses on the total asset portfolio and specific provisions relate to anticipated future losses on a specific loan or asset.
Data on both general and specific provision charges (by sector), made by UK banks and building societies, are captured by the Bank’s quarterly profit and loss return (Form PL).
Arrears: A loan is considered to be in arrears if the borrower has missed one or more of their obligated loan repayments.
The Bank, in conjunction with the Financial Conduct Authority, publishes the Mortgage Lending and Administrators Return (MLAR) data. This captures the share of mortgage lending in arrears. It also provides detail on the severity of these arrears, indicated by the value of missed payments relative to the total outstanding mortgage balance. In addition, it contains information on the share of these loans that are receiving forbearance, and the share of new arrears cases. Arrears reported in the MLAR data relate only to loans where the amount of arrears is 1.5% or more of the borrower’s current loan balance.
Possessions: A loan is considered to be ‘in possession’ if the lender takes the property against which it is secured against into their possession.
The MLAR data capture the number and value of mortgages which flow into and out of ‘possession’ each quarter, and the stock of mortgages against possessed properties which remain on lenders’ balance sheets.
Write-offs: Write-offs occur when a loan is considered uncollectible and the loan amount is deducted from the firm’s recorded assets.
The Bank publishes data on the nominal value of assets that have been written off (net of any written off debt that is later ‘recovered’) by UK banks and building societies each quarter. These data are collected on Form WO, which aims to capture any changes in the value of assets over the quarter that do not result from market valuation changes.
Comparison of data collections
Depending on the context of interest, users can select the indicator most suitable for their purpose. In addition to the indicator type, the differing data sources have a number of other features that users should bear in mind – including differences in the sectors by which the data are broken down, the coverage, timeliness and history of the data. These features are explained in this section, and summarised in Table 1.
Due to the differing data sources and purposes, each indicator offers a different breakdown of the data by sector (see Figure 1).
MLAR data focuses on secured lending, offering a breakdown of performance between regulated and non-regulated, and securitised and unsecuritised mortgages. Regulated mortgages can broadly be defined as those to owner occupiers. Non-regulated lending is mostly comprised of buy-to-let lending, though includes some other types of mortgages, including owner-occupier loans that were taken out before the relevant regulation came into effect.
Other data collections offer a wider breakdown of the performance of loans and other assets by sector. There are several sector breakdowns that write-offs and provisions data have in common. These include private non-financial corporations (PNFCs), other financial corporations (OFCs), households (HHs) and non-residents. However, there are some differences in definitions that mean these data are not directly comparable. Further detail of definitions can be found in the MLAR guidance notes and the statistical form definitions.
The MLAR data has full coverage of the regulated mortgage market, covering all firms that are authorised to undertake and administer regulated mortgage lending. The unregulated lending by these firms is also captured. The firms include banks and building societies as well as other authorised lenders. Activity by lenders with only non-regulated lending is not captured.
For write-offs and provisions, data are collected to capture the majority of the market. For provisions on Form PL, the current coverage of the quarterly collection is approximately 95% of the stock of provisions. For write-offs on Form WO, the current coverage is approximately 70% of total lending by UK MFIs to other UK residents. Grossing up techniques are applied by the Bank to estimate the remainder of data for non-reporting institutions. This enables estimates of data for the whole market to be produced and published.
Frequency/ reporting timeline
All three data sets are collected quarterly. The data on write-offs are published eight to nine weeks after quarter-end, whilst arrears and possessions data are published approximately ten weeks after quarter-end. The quarterly provisions data is published annually, approximately seven months after year-end.
Series start date
The arrears and possessions data form part of a more recent collection, having only been collected and published since Q1 2007. For provisions, a slightly longer-time series is available with the data starting from Q1 2004. The write-offs data introduced here have been collected since Q1 2008, however prior to this some write-offs data were collected quarterly on a slightly different basis.
With the introduction of the new International Financial Reporting Standard (IFRS) 9, a new expected-loss impairment model for loans has been introduced. Under the previous model, lenders were only required to recognise credit losses, and hence make a provision, when evidence of a loss was apparent, for example through a missed payment. Under the new ‘expected credit loss’ framework banks are required to recognise potential losses at all times, taking into account past events and current and projected economic conditions. Under the previous model, lenders would provision for a portion of all loans extended regardless of individual risk profiles. This was known as a ‘general provision’. Under the new framework, lenders should assess the risks of assets individually, so all provisions should be ‘specific provisions’.
The more timely recognition of bad and doubtful debts is expected to cause an increase in the total amount of provisions when reporting institutions adopt IFRS 9. For Form PL this change took effect from 2018 Q1. It is also expected that there will be a corresponding increase in the release of provisions as debt obligations are met and provisions are unwound.
The Bank of England publishes several series that serve as indicators of asset and loan performance. There are a number of common and exclusive sector splits offered within the data sets and each indicator captures a different view of asset performance. Users’ needs will determine which data are most appropriate for their specific requirements.
We would be pleased to receive views from users on the usefulness of the data sets or other potential developments. Please send any comments or queries to email@example.com.