DP1/26 – Future banking data

Privacy statement

By responding to this discussion paper, you provide personal data to the Bank of England. This may include your name, contact details (including, if provided, details of the organisation you work for), and opinions or details offered in the response itself.

The response will be assessed to inform our work as a regulator and central bank, both in the public interest and in the exercise of our official authority. We may use your details to contact you to clarify any aspects of your response.

The discussion paper will explain if responses will be shared with other organisations (for example, the Financial Conduct Authority). If this is the case, the other organisation will also review the responses and may also contact you to clarify aspects of your response. We will retain all responses for the period that is relevant to supporting ongoing regulatory policy developments and reviews. However, all personal data will be redacted from the responses within five years of receipt. Find out more about how we deal with your personal data, your rights or to get in touch.

Information provided in response to this paper, including personal information, may be subject to publication or disclosure to other parties in accordance with access to information regimes including under the Freedom of Information Act 2000 or data protection legislation, or as otherwise required by law or in discharge of the Bank’s functions.

Please indicate if you regard all, or some of, the information you provide as confidential. If the Bank of England receives a request for disclosure of this information, we will take your indication(s) into account, but cannot give an assurance that confidentiality can be maintained in all circumstances. An automatic confidentiality disclaimer generated by your IT system on emails will not, of itself, be regarded as binding on the Bank of England.

Responses are requested by Tuesday 5 May 2026.

The PRA prefers responses to be sent via email to: DP1_26@bankofengland.co.uk.

Alternatively, please address any comments or enquiries to:
Future Banking Data
Prudential Regulation Authority
20 Moorgate
London
EC2R 6DA

Executive summary

The Prudential Regulation Authority (PRA) collects a range of regulatory data from firms to support the delivery of its statutory objectives. These data are required for the PRA to be able to supervise firms effectively, assess risks to financial stability, and inform policy development and evaluation. To support these functions, the PRA needs timely, high-quality, and relevant data. However, producing such data can be costly for firms, making it essential that reporting requirements are proportionate.

The PRA is reviewing its strategic approach to regulatory reporting for banks^{footnote [1]} through the Future Banking Data (FBD) programme. The FBD programme aims to deliver tangible cost reduction for banks in line with the PRA’s secondary competitiveness and growth objective, as well as improvements to the relevance, quality and timeliness of data collection.

In this discussion paper (DP), the PRA sets out its latest thinking on banking data collections with a view to pursuing a programme of pragmatic and incremental changes over the coming years. The PRA is seeking views from firms to help shape this programme.

The DP sets out how the PRA uses data for supervisory, operational, and policy purposes including supporting risk assessment, to confirm compliance with its rules, for peer group analysis, for financial stability monitoring, and for policy analysis, research, and evaluation. The PRA uses a wide range of regulatory data covering capital and liquidity, credit quality, market and counterparty risk and many other aspects of firm risk and performance. Regulatory data are supplemented by management information provided by firms – which is central to building the PRA’s understanding of how firms are taking decisions – and ad-hoc data collections that focus on new and emerging issues.

The PRA recognises that data provision can represent a substantial cost for firms. Balancing this, the information provided to supervisors is also highly valuable, contributing significantly to maintaining the safety and soundness of firms, supporting the PRA’s ability to respond to events and more broadly helping protect the UK economy against damaging financial crises. Data also support the efficient and evidence-based design of policy, helping the PRA deliver regulatory outcomes in a cost-effective manner, consistent with its secondary competitiveness and growth objective.

While the PRA’s data collection approach already aims to be proportionate, there are areas which could be further streamlined and modernised. Potential areas for improvement include reviewing what data it collects, from which firms, and at what frequencies to identify further efficiency opportunities; evaluating whether legacy collection processes reflect current best practices; making instructions clearer and easier to find; and checking for, and if need be, closing, data gaps in particular around new and emerging risks. The PRA hopes to prioritise its work across and within each of these areas by improving its understanding of the drivers of data costs for firms.

This DP suggests a direction for incremental reform. This would include extending the initial template deletions the PRA recently implemented by reviewing a wider range of collections. This work would also draw on the approach followed for reporting as part of the PRA’s Strong and Simple initiative, which has already simplified liquidity returns for Small Domestic Deposit Takers (SDDTs), and which will deliver simplifications to capital reporting for SDDTs from 1 January 2027. Significant streamlining and reductions in reporting were already implemented for insurers as part of the Solvency UK reforms which concluded in 2024.

To guide the FBD programme’s work this DP suggests four broad principles, which build on current PRA practice: (i) anchoring the data the PRA collects in its objectives, (ii) collecting data ‘once and well’ by minimising the data collected to meet its objectives but maximising its use, (iii) making it easier for firms to supply data, and (iv) ensuring the data collected remain fit for purpose over time. Proportionality and cost-benefit assessment are central within these principles, consistent with the PRA’s intention to collect only high-value data at an appropriate cost.

In making reforms to reporting, the PRA would seek to balance a number of trade-offs, including the extent of data model standardisation (including definitions), cross-firm comparability, international alignment, granular data vs aggregated formats, and the role of and relative balance between regular reporting and one-off requests. There is also a balance between pursuing longer-term holistic developments and delivering initial improvements more quickly.

In light of responses to this DP, the PRA will work with firms to develop a roadmap for pragmatic and incremental reforms.

Scope

This DP is relevant to PRA-authorised UK banks, building societies, PRA-designated UK investment firms, and their qualifying parent undertakings, which for this purpose comprise financial holding companies and mixed financial holding companies, as well as credit institutions, investment firms, and financial institutions that are subsidiaries of these firms, regardless of their location. It is not relevant to credit unions.

The FBD programme is a PRA-led initiative focussing on the data collected by the PRA, including regulatory returns set out in the PRA Rulebook, non-Rulebook collections, and data provided by firms in the course of PRA supervision. Other data collected by the Bank, such as that collected for SMF operations or for statistical purposes, and data collected by other authorities such as the FCA are described for context and are not within the scope of this DP.

The FBD programme is separate from, but joined up with, the recently announced review of the UK’s statutory transaction reporting regimes.^{footnote [2]}

Responses and next steps

This DP closes on Tuesday 5 May 2026. The PRA invites feedback on the topics discussed in this DP. Please address any comments or enquiries to DP1_26@bankofengland.co.uk.

1: The role of banking data at the PRA

1.1 Data are fundamental to the PRA’s ability to deliver its statutory objectives. Data underpin effective supervision, support work on the stability of the UK financial system, and inform decision-making across a wide range of policy and operational areas.  

1.2 The PRA’s unique remit shapes the scope, diversity, form, and timeliness of the data it needs. This chapter sets out how the PRA uses data to advance its objectives and carry out its functions.

Objectives and functions of the PRA

1.3 The PRA’s primary objective relating to banks is to promote their safety and soundness.^{footnote [3]} The PRA also has two secondary objectives, namely facilitating effective competition in the markets for the firms it regulates; and facilitating the competitiveness of the UK economy and its growth in the medium to long term.

1.4 In pursuing its objectives the PRA also has regard for aspects of the Government’s economic policy, as set out in a letter from HM Treasury (HMT) to the Prudential Regulation Committee (PRC). These aspects include the contribution of the financial services sector to economic growth, proportionate and effective regulation, international competitiveness, and responsible risk-taking, alongside maintaining financial stability.

The PRA’s supervisory approach and resulting data needs

1.5 The PRA delivers its objectives through responsive, risk-based supervision, and by developing regulatory standards and policies that set out the PRA’s expectations of firms. The PRA’s approach to supervision is forward-looking, judgement-based, and focused on the issues and firms that pose the greatest risk to the stability of the UK financial system, as set out in the PRA's approach to banking supervision. The PRA’s approach to policymaking has the aim of being a strong, accountable, responsive and accessible policymaker as set out in its approach to policy document.

1.6 Day-to-day supervision of individual firms is central to promoting their safety and soundness. The PRA’s approach is centred on firm engagement and setting clear expectations. Supervisors’ prioritisation of activities and decision-making is informed throughout by firm supplied data, including regulatory returns set out in the PRA Rulebook and non-Rulebook collections, illustrated in Chart 1, plus management information (MI), statutory transaction reporting collections, publicly available data sets, and qualitative inputs. The use of these data is set out in Use case 1: Firm supervision, including the role of data in the Proactive Intervention Framework, and its use in Periodic Summary Meetings and for peer benchmarking.

1.7 The PRA applies a risk-based and judgement-led supervisory model which involves:

• Proactive risk identification: Supervisors assess the risks firms pose to the PRA’s objectives, drawing on both firm-provided data and scenario analysis. 
• Proportionality: The intensity of supervision is scaled to the potential impact of a firm’s failure, with greater scrutiny for systemically important institutions, reflected in the different volumes of data collected from firms, shown in Chart 2. 
• Supervisory judgement: Quantitative data are combined with qualitative insights to form forward-looking assessments of resilience, both at a firm level and across the system. 
• Supervisory Review and Evaluation Process (SREP): Supervisors assess whether firms are adequately capitalised and liquid given their risk profile, integrating institution-specific evaluations with thematic reviews and stress testing. All firms must demonstrate effective risk identification and resilience planning. 
• Role of firm management information: Understanding how firms use their own management information is critical to assessing governance and decision-making. 

1.8 Regular and sufficiently detailed data support proactive, prioritised and risk-focused supervisory decisions which can save resources for both the PRA and firms. In many cases, the data requested by the PRA are also used by firms in the course of running their businesses, managing their risks, or confirming compliance with regulations.

Chart 1: PRA banking data, 2024

Datapoints collected across regulatory, statistical, and stress test returns

Footnotes

• Count represents datapoints for PRA regulatory returns (PRA 110, COREP, FINREP, PRA Other), Statistical, STDF and Resolution data collection for the 2024 period. Figures exclude zeros and use the most recent submission. Data on the statutory transaction reporting regimes in the UK (EMIR TR, SFTR, and MiFID2) are not included in this and subsequent charts as they are out of scope of the FBD programme (see Footnote 2).

Chart 2: Larger firms submit proportionately more data

Number of datapoints submitted annually, by firm size

Footnotes

• Firm size measured by total assets: Small (<£20 billion), Medium (£20 billion–£250 billion), Large (>£250 billion). Historic data do not reflect forthcoming reductions in reporting for SDDTs under the Strong and Simple framework.
• Average datapoints are calculated by summing all datapoints submitted by reporting entities within each firm structure, grouped by size, then dividing by the number of firm structures in each size category. Datapoints are from regulatory return submissions for the 2024 reporting period. Figures exclude zeros and use the most recent submissions even if resubmitted after 2024.

Crisis supervision versus business-as-usual needs

1.9 The PRA’s crisis framework, consistent with the Authorities’ Response Framework used by the Bank (including the PRA), FCA, and HMT, is designed to enable rapid response to destabilising events. In periods of financial stress, timely data are essential to assess resolvability and systemic risk. 

1.10 The data requested by the PRA will vary by the type of crisis event and sometimes by type or size of firm. Regardless of these variations, in most circumstances the most critical factor is that firms have the capability to provide relevant, timely, accurate data under pressure. This information supports decisions that can: 

• Maintain the soundness of firms, the banking sector, and the wider financial system;
• Protect deposit holders; 
• Enable orderly resolution or recovery; and 
• Reduce the likelihood of government or taxpayer intervention. 

1.11 In a crisis, the PRA may prioritise timeliness over comparability or consistency of data across firms, whereas in business-as-usual conditions, comparability will often take precedence to ensure consistent analysis across firms. The box Use case 2: Crisis supervision provides examples.

Cross-sectional analysis and policy needs

1.12 System-wide and cross-firm analysis supports risk identification and investigation, policy development including cost-benefit analysis (CBA), research, financial stability monitoring, thematic reviews, and can inform supervisory resource allocation. As outlined in Use case 3: Cross-sectional analysis, many of the data sources used for firm supervision are also used in this work. Non-Rulebook submissions (sometimes called ‘ad hoc’ requests) can be particularly valuable for exploring emerging risks, allowing supervisors to better understand and monitor specific issues facing certain cohorts of firms. Other data sources including commercially available datasets are also used to support these analyses. Historically, cross-firm analysis has been focused on quantitative inputs, specifically on firms’ regulatory returns. However, new analytical techniques are permitting further analysis of qualitative data and firm management information.

1.13 Policy design is particularly data-intensive, particularly the calibration, impact analysis, and CBA phases. Similarly, research, including analyses specifically intended to evaluate policies that have been implemented, can generate very particular data demands, often requiring consistency in data over time to assess outcomes before and after implementation, or construct econometric comparisons against non-implementation counterfactuals. These data needs are explored in Use case 4: Policy and research, which also highlights bespoke databases and models the PRA maintains, based on cleaned firm data from regulatory returns, which are a reusable resource used to support a range of policy impact analysis and research.

Diversity of PRA data collection

1.14 The PRA’s use of data has evolved over time, shaped by changes to its remit and responsibilities, evolving understanding and the shifting landscape of risk, industry changes, innovations in markets and products, and international developments. Consequently, the prudential data estate has evolved significantly, from the origins of the current framework under the Financial Services Authority (FSA) to the present.

1.15 Notable changes have arisen from the implementation of Basel II and III, the development of EU data collection standards under the Capital Requirements Regulation – specifically Common Reporting (COREP) and Financial Reporting (FINREP) – and, both prior to and following the UK’s exit from the EU, the introduction of specific UK collections to address gaps in the EU framework, including for stress testing.^{footnote [5]}

1.16 Reflecting the breadth of the PRA’s responsibilities, individual data collections can serve multiple purposes and are defined in different ways. 

1.17 The scope of collections is broad, including but not limited to: capital adequacy, liquidity, credit quality and credit risk, market risk, and counterparty risk, along with other factors such as large exposures, as illustrated in Charts 3 and 4. Measured by data points, liquidity data make up the most significant proportion of the PRA’s collections, reflecting banks’ role in conducting maturity transformation and the PRA’s focus on effective supervision of liquidity through understanding firms’ outflows and inflows, and their capacity to balance these.

1.18 The majority of regular data collections are specified in PRA Rules. However, several annual collections – most notably the data collected to support the Bank’s stress testing programme (described further in Use case 5: Focus on the Stress Test Data Framework) are not.

Chart 3: Banking data cover a large variety of themes

Regulatory datapoints by reporting theme (excludes liquidity)

Footnotes

• Regulatory returns are grouped under a primary theme for the purposes of illustration; some data supports analysis on multiple themes. Liquidity regulatory returns are excluded. Datapoints are from regulatory return submissions for the 2024 reporting period. Figures exclude zeros and use the most recent submissions.

Chart 4: The composition of the data PRA collects reflects the risks firms face and pose

Regulatory datapoints by reporting theme

Footnotes

• Firm size measured by total assets: Small (<£20 billion), Medium (£20 billion–£250 billion), Large (>£250 billion). Historic data do not reflect forthcoming reductions in reporting for SDDTs under the Strong and Simple framework.
• Average datapoints are calculated by summing all datapoints submitted by reporting entities within each firm structure, grouped by size, then dividing by the number of firm structures in each size category. Datapoints are from regulatory return submissions for the 2024 reporting period. Figures exclude zeros and use the most recent submissions even if resubmitted after 2024.

The PRA’s powers to collect data

1.19 The PRA’s principal mechanism for collecting data from firms is through rules it can make under section 137G Financial Services and Markets Act 2000 (FSMA). This power allows the PRA to make rules it considers necessary or expedient for advancing its primary objectives, while also pursuing its secondary objectives and considering its various statutory ‘have regards’. These rules are set out in the PRA Rulebook.

1.20 The PRA also has power to make non-rules-based requests for data, for example under s.165 FSMA, where the PRA has power to collect data that supports the exercise of its FSMA functions, or is reasonably required by the Bank in pursuance of its financial stability objective. Additionally, the PRA frequently makes data requests of firms on a voluntary basis as part of ongoing supervisory engagement. The Bank (ex PRA) collects data under other provisions, including through statutory powers under the Bank of England Act 1998, and under the contractual arrangements for the Sterling Monetary Framework (SMF).

Data sharing with the Bank and FCA

1.21 Much of the data collected by the PRA and the Bank will be confidential and subject to restrictions on disclosure, including under section 348 FSMA and the Bank of England Act 1998. This includes restrictions on sharing between the Bank, PRA and FCA without the consent of the firm that supplied the data (or, if different, the person to whom it relates) or unless a legal gateway is available. Legal gateways require analysis of the type of data to be shared and the functions of the recipient. Information sharing and the protection of confidential information is an important part of the Memorandum of Understanding between the Bank, FCA, Payment Systems Regulator, and the PRA, made under section 99 of the Financial Services (Banking Reform) Act 2013.

1.22 The PRA makes use of data provided by the FCA, such as the Product Sales Data for mortgages, and MiFID2 transaction reports. The PRA and FCA also have some shared collections; in most cases the collection itself is made by the FCA, for example some of the major mortgage datasets described in Use case 6: Mortgage data collections.

International engagement and policy development

1.23 In addition to data collected for the PRA and Bank’s policy and monitoring, the PRA also collects data to support the UK’s international engagement, for example at the Basel Committee on Banking Supervision (BCBS) and Bank for International Settlements (BIS). Impact studies based on UK banks’ data support the PRA’s negotiations to ensure global standards are fit for the UK as a global financial centre and, where appropriate, are relevant for domestic UK banks.

1.24 The Bank also shares some statistics with other public bodies such as the ONS and international organisations including the Organisation for Economic Co-operation and Development (OECD) and International Monetary Fund (IMF) to support macroeconomic monitoring and analysis of the global financial system.

2: Recognising the challenges

2.1 Delivering the PRA’s statutory objectives requires timely, high-quality, relevant and sufficiently granular data in order to supervise firms and support financial stability, as shown in the case studies. Sourcing, transforming, validating, and submitting the data required for robust supervision and policy work will inevitably entail a cost on firms. The aim of the FBD programme is to deliver tangible cost reduction in banking regulatory reporting in line with the PRA’s secondary competitiveness and growth objective, as well as improvements to the relevance, quality and timeliness of the data the PRA collects which will further support its primary objective of safety and soundness.

2.2 The PRA has identified four areas which might benefit from streamlining, modernisation or other improvements as noted in this paper. First, the PRA’s data estate represents the accumulation over time of multiple data collections serving different requirements. It is to be expected that the supervisory and regulatory value of some of these collections will have evolved and some collections may now generate less value than was intended when they were designed. Second, there is potential to make collection processes clearer and more coherent, including potentially across UK authorities. Third, firm feedback suggests that PRA reporting instructions could be made clearer. And fourth, new and revised data are needed in some areas to support mitigation of new and emerging risks or to close known gaps.

The PRA’s data estate includes multiple collections designed for multiple purposes

2.3 The PRA’s data collection from banks spans over 400 templates of varying sizes and complexity. As an example, capital is covered in one form or another in around 120 of these templates, including within COREP, FINREP, legacy FSA templates, PRA templates, and the STDF. Similarly, across the PRA, wider Bank, and data received from the FCA, credit risk is covered in over 100 different collection templates.

2.4 The range of these templates reflects changes in bank regulation in the UK over time and the evolution of statistical collections. They include Bank of England statistical data that have been collected for decades, FSA templates designed prior to the formation of the PRA, EU collections designed after the global financial crisis (now adopted into PRA Rules following EU exit), and data fields which are specific to the UK market, or, as with STDF, support new initiatives not then envisaged within EU legislation.

2.5 Given the differing institutional and legal origins of these data collections, it may not have possible or cost effective at the time to design newer templates taking other collections into account, even though some of these collections focus on similar risks. Although explicit duplication across templates is relatively uncommon, collections sometimes ask for similar underlying data to be cut or aggregated in different ways.

2.6 There are certain legacy data collection templates whose ongoing supervisory value may have reduced or been superseded by new reporting or given events since their creation. The PRA has already begun a work programme to address these, with the first phase of deletions, mostly of FINREP templates, implemented in December 2025.

2.7 When designing any new collection, the PRA evaluates a range of factors to determine the minimum necessary group of firms within scope and the relevant consolidation levels, and the appropriate frequency, so that the cost of the collection is proportionate to the underlying risks. In some cases, however, EU rules which applied when the UK was a member have constrained the PRA’s ability to tailor collections to more closely reflect the characteristics of the UK banking sector. Given changes to the risk environment, new methods for obtaining information, updated supervisory and policy priorities, and more flexibility for UK regulators, there may be scope to reevaluate who submits what, when, including potentially revisiting materiality thresholds.

2.8 The PRA has already used its post-EU withdrawal flexibility to deliver substantial reductions in reporting for insurers, as noted earlier. It has also increased proportionality in reporting as part of introducing the Strong and Simple framework. SDDTs have benefited from simplified liquidity reporting since 2024, and a simplified suite of capital-related reporting has been designed to ensure small firms conducting simpler business will only have to submit a more limited set of data, tailored to their circumstances.^{footnote [9]}

Potential to make collection processes more coherent

2.9 Separate to what data the PRA collects, there is potentially also scope to modernise how the PRA collects data. The PRA and the Bank have three main collection methods: the Bank of England Electronic Data Submission (BEEDS), an online portal run by the Bank; RegData, the FCA’s data collection platform, used for many PRA collections; and submission of spreadsheets by email (the most common case for non-Rulebook collections). Each of the collection portals receives templates in a range of different formats, but firms must supply each template in a predetermined format. Most RegData collections are encoded in XBRL, with some in XML; BEEDS templates are in a mix of XBRL, XML, and Excel spreadsheets.

2.10 Firms submit data continually throughout the year, with peaks after the end of each quarter (Chart 5). Remittance dates are sometimes set in business days, with calendar days for other collections. Looking only at the templates currently submitted monthly, some firms would have deadlines on 15 different dates measured in business days (from 7 days to 35 days), another at 20 calendar days, and one more expressed as ‘1 month’. Feedback from firms has suggested that reviewing the alignment and coordination of collection dates could reduce challenges around peak submission periods.

2.11 The PRA recognises that, given change costs (discussed further below), these differences may not be issues that merit fixes in their own right. Nonetheless, in taking forward reforms, the PRA would plan to consider the formats and collection methods that work best, including timing considered holistically with firms’ other data submissions. Fundamental revisions to technology platforms for submitting data are not planned within the scope of FBD at present.

2.12 Finally, for repeated non-Rulebook collections, there may be benefits – including through clarity to firms to support systems investments – to incorporating these in rules and collecting via core reporting systems.

Chart 5: Firms submit data throughout the year

Spikes reflect weekly liquidity submissions and quarter-end data

Footnotes

• Datapoints are regulatory return submissions received in 2024 reporting period.

Instructions could be clearer

2.13 Feedback from firms has indicated there is scope to improve the clarity of the PRA’s reporting requirements. Three parts of the Rulebook directly address banking data collection: Regulatory Reporting, Reporting (CRR), and Reporting Pillar 2.^{footnote [10]} However, numerous other parts of the Rulebook also require reporting related to specific policies, for example Large Exposures (CRR) which, in some cases, reflects rules developed when the UK was a member of the EU that are now part of the PRA Rulebook.

2.14 For data collections developed by the PRA, substantial efforts are made to ensure instructions are clear and readily understandable by PRA-supervised firms, with guidance that is updated in response to queries. Looking across the whole data estate, however, the PRA recognises firms can face challenges in identifying what to submit. As well as requirements sitting in different Rulebook parts, reporting on individual topics is also determined by at least thirty different criteria relating to firm size, although these represent only around a dozen distinct thresholds.

2.15 Reporting templates are often provided in Excel format as appendices to the relevant rules, typically with distinct templates for firms using different accounting standards. Instructions are often appended to the rules, and are also shared on the PRA’s website, as are the XBRL taxonomies, where applicable. Instructions and templates inherited from the EU can still contain legacy references to EU rules, regulations, and guidance meaning firms need to find the PRA’s own instructions manually.

2.16 As well as increasing the burden of providing data, insufficiently clear reporting instructions can result in firms reporting data that is not required and can also have a negative impact on data quality in general, creating challenges for both supervisors and firms themselves. There is scope to consider whether other approaches for specifying the data to be submitted, such as data dictionaries, can provide benefits.

Data gaps have emerged since some templates were designed

2.17 The PRA’s data estate has been continuously amended to reflect policy changes and emerging risks over time. Nonetheless, known gaps remain where new or different data are required but are not presently available. One specific area of focus for the PRA and Bank is risk in the non-bank financial institution sector.

2.18 Timeliness of data is important in all contexts but can also be a particular driver of gaps data gaps and supervisory or financial stability blind spots if not calibrated correctly. Given a risk environment that evolves more quickly than before due to technological change, for some specific themes data frequencies may need to be increased or more timely submission required.

There is scope to model costs and benefits better

2.19 The PRA recognises that, while reporting improvements might provide substantial ongoing long-term cost reductions for banks, these changes are also likely to represent short-term implementation costs for firms. Therefore, it is vital that the PRA understands both the costs of ongoing reporting and the costs of different types of change to calibrate individual reforms and build a roadmap for reform.

2.20 The cost of preparing data can depend on factors such as data volume, complexity, granularity, availability and the number and novelty of the data breakdowns requested. The PRA’s current modelling of firm costs is based on a firm survey conducted in 2023 to support the implementation of Basel 3.1. Given a relatively low response rate by firms, the evidence base on costs lacks detail and may not adequately reflect all relevant elements.

3: Practical and feasible steps for reform

3.1 As already noted, the PRA has already begun reforms to its banking data collections, with reduced reporting for SDDTs under the Strong and Simple framework, and under FBD, an initial round of whole template deletions, implemented with effect from 31 December 2025.

3.2 This chapter sets out a potential approach for progressing FBD with the aim of delivering further burden reduction for firms, recognising the challenges outlined in Chapter 2. That approach is based around four principles which guide decisions regarding the data the PRA collects from banks. Many of those decisions, including on the sequencing of reforms, require competing factors to be balanced, with the most significant trade-offs outlined below.

Principles for Future banking data

3.3 As outlined in Chapter 1 and the Use cases in this DP, data are a critical input to the PRA’s activities to advance its objectives, and likewise for the Bank as a whole. It is important that the data the PRA collects are targeted and proportionate, their purposes transparently articulated, and collection is efficient and timely. Changes – both reforms and ongoing maintenance – must be designed and executed in a measured and cost-effective manner. These considerations underpin the four suggested principles for FBD, which build on current PRA practice.

Principle 1: Our data collection is objectives-driven

• The data we collect is driven by the PRA’s statutory objectives, or, within the scope of the PRA's legal powers, the wider Bank’s financial stability objective.
• We link collections explicitly to specific supervision and policy responsibilities, supported by robust business cases and CBA.
• The nature of our collections reflects the use case for the data – sometimes we have a temporary need for data, sometimes longstanding; sometimes our focus is on specific firms, at other times we will need a cross-firm view as part of peer group or cross-sector analysis, policy, and research.
• By focussing carefully on the costs and benefits of collecting data, our collections are proportionate to the risks firms pose and face, and also support the PRA’s secondary competition objective and secondary competitiveness and growth objective (SCO and SCGO).

Principle 2: We aim to collect data from a firm ‘once and well’

• Wherever possible, we avoid duplication and ensure definitions are coherent across collections, supporting proportionality.
• We use individual collections to their fullest extent, including by considering alternative sources and analytical methods in all collection design.
• We will improve – subject to legal restrictions on powers of data collection and information sharing – our governance and processes (including metadata) to better enable data reuse across multiple responsibilities of the Bank, PRA, and other authorities.

Principle 3: We prioritise making it easier for firms to supply high-quality data

• Our templates, rules, reporting instructions, and other supporting guidance are easily navigable to firms and their staff, including at smaller firms.
• Underlying data definitions are clear and specific, using dictionaries to support consistent reporting of data points.
• We make it easy for firms to engage with us on data interpretation questions when we consult on instructions.
• Data submission is as straightforward as possible, with options, as appropriate, to suit a range of firm circumstances.

Principle 4: Our data collection must remain fit for its purpose over time

• We adapt the data we collect as our needs and circumstances change, to ensure principles 1-3 continue to be met, while recognising the costs of change, subject to CBA.
• To do so, we review policy needs and business cases, using sunset provisions or pre-agreed assessment points as appropriate.
• We design our processes to enable changes to be made by firms as simply and cheaply as possible, including by highlighting where legislative changes can support efficiency.

3.4 The substance in these principles is not new: for example, all rules made by the PRA, including those which require reporting are underpinned by its objectives and supported by CBA as set out clearly in the relevant consultation papers and policy statements. Similarly, non-Rulebook requests for data are subject to internal governance which includes cost and proportionality considerations. Rather, by setting out these principles for discussion in this paper, the PRA hopes to increase understanding of its approach to data and reporting and to be transparent about how it intends to deliver any future phases of the FBD programme.

3.5 Principle 1 embeds the PRA’s objectives within FBD, emphasising that the overall data ‘ask’ is commensurate with, and relevant to, supervisory and policy decision-making needs, and that this is transparently communicated. This principle supports data collection that is proportionate by both risk and firm impact. Under Principle 2, this ‘ask’ is joined up – to the extent possible legally and technologically – across authorities, with requests structured in a coherent way that enables efficient reuse. Together, these two principles point to a rightsized data estate that is likely to be somewhat smaller and simpler than today’s, with the same or better support for the PRA’s general (safety and soundness) objective but at tangibly lower cost for firms.

3.6 Making data easier to prepare and provide is another major way to reduce firms’ costs, as recognised in Principle 3. This includes being clear about what data are to be provided and by whom, a concern particularly relevant to smaller and mid-size firms, and with implications for data quality for all firms. Principle 3 is relevant both to rules-based and non-Rulebook collections, where data dictionaries and template formats may be able to reduce how much data processing is needed for non-Rulebook collections.

3.7 Finally, Principle 4 recognises that the PRA’s data needs will never be static. The risk environment will always continue to evolve, and as firms’ business models evolve, policy and supervisory priorities will adapt. This means that new data items will always be needed – at present, for example, there are material gaps in our understanding of the risks posed by non-bank financial institutions. But, under Principle 4, regular reviews will become the norm in considering if some data has become less relevant and can be limited, deprioritised, or retired. Principle 4 applies both in the new steady state, but also on the path towards it: this principle, along with the robust alignment with objectives and focus on costs under Principle 1, will guide the sequencing of PRA reforms.

3.8 These ‘common sense’ principles will guide the PRA’s transition towards a more efficient data estate and maintaining it effectively.^{footnote [11]}

Navigating trade-offs

3.9 Every data collection has different characteristics, and in designing collections to meet the four principles above, a range of trade-offs must be weighed up.

3.10 Timeliness vs comparability – the PRA’s core regulatory collections provide data that is directly comparable across firms but supplied mainly with a lag (PRA110 is a notable exception to this). In urgent circumstances, the appropriate balance can switch, as discussed in Use case 2: Crisis supervision, although accuracy would always be required. Firms’ own management information can provide data with a lower lag but is frequently hard to reconcile and compare across firms. Therefore, in the present reporting environment, this presents a trade-off, although there may be potential for greater consistency in data definitions and/or dictionaries to make both aims jointly achievable, as well as simplifying the creation of new data at short notice in unforeseen circumstances.

3.11 Standardisation vs flexibility – Standardising definitions across multiple data collections presents a clear direction for reducing costs and would support de-duplication. Naturally, that standardisation would have to work for the multiple purposes of the PRA and the Bank and in some cases across other regulators.

3.12 The PRA is open-minded to the possibility that further advances in machine learning or other forms of AI might mitigate the costs for the PRA associated with receiving less consistent or less structured data, including data that is defined by different firms in different ways.

3.13 Aggregate vs granular – the more granular a collection is, the greater the potential reuse it may have. This is particularly relevant to UK mortgage data, as outlined in Use case 6: Focus on mortgage collections. Nonetheless, granular data may carry additional costs if firms must merge data items across multiple systems or if transformations/enrichments are needed to produce a granular return. It may also be harder to link granular data elements to the PRA’s statutory objectives and firms may wish to be able to readily replicate the reporting metrics computed by the PRA, which could be another source of cost. The appropriate extent of granular collection will depend on the costs and the benefits to both firms and the PRA.

3.14 Regular vs ad hoc – At present, there are some differences between the PRA’s approach for regular requests to firms, the majority of which are codified in the PRA Rulebook, and that for non-Rulebook requests. The two approaches differ in the design process, instructions, and template formats. Typically, one-off requests are relatively self-contained, whereas Rulebook collections often have separate but interlinked rules, templates, instructions, and taxonomies.

3.15 Rulebook and non-Rulebook returns present different cost characteristics, reflecting the need for firms to develop reporting systems to support Rulebook collections versus the often manual processing for one-off requests. However, supervisors are often able to provide firms with more flexibility around the data provided and submission timing for one-off requests.

3.16 This said, consistent with the suggested Principles 2 and 3, improvements to the clarity of data requests and instructions, specific data definitions, potential use of data dictionaries, and more standardisation of formats should reduce the costs of both regulatory and non-Rulebook collections.

3.17 International alignment vs UK tailoring – as noted above, there remain a number of areas where the PRA has inherited reporting requirements from the EU which may present opportunities for cost reductions for firms operating in the UK. This is particularly likely where synergies can be found across PRA and other Bank or FCA data collections. Consistent with Principle 2, the PRA would seek to maximise alignment across UK authorities where possible recognising different remits and objectives.

3.18 However, the PRA recognises that having data collections align to those of overseas authorities can mean lower aggregate costs for firms submitting the same or similar collections in multiple jurisdictions, such as the EU.

3.19 Under the FBD programme, the PRA would seek to balance these factors where possible. International alignment is not an end in itself but is a key factor in delivering the PRA’s secondary competitiveness and growth objective and does reflect the UK’s standing as a global financial centre.

3.20 Data continuity vs decommissioning – consistency in data series across time is a material asset for certain types of analysis and is particularly advantageous for research and policy development and evaluation, as described in Use case 4: Policy and Research. It is also typically less costly for firms to continue to provide data once the start-up costs have been absorbed. This does not, however, mean that data collections, once established, should continue in perpetuity if their utility has waned.

3.21 For any given element of data collected, the pros and cons of decommissioning would be unique. When considering whether to retire collections, the PRA considers on a case-by-case basis whether its objectives might be advanced using alternative means at a lower cost. In the future, those alternatives could increasingly include synthesising legacy data series via machine learning or econometric interpolation of other sources.

3.22 Minimising rework vs delivering improvements sooner – the PRA recognises that reporting systems represent a material investment cost to firms and that certainty is important in providing banks the confidence to take sound investment decisions.

3.23 Given the scope of the PRA’s banking data and interlinkages with the Bank and FCA’s responsibilities, the PRA considers a ‘big bang’ approach – attempting a reform of the entirety of the data estate at one time – would present excessive up-front costs and delivery risks. A road map that accepts some amount of ‘rework’ may enable firms to gain some benefits sooner, for example by simplifying some templates on a legacy system before migrating them to a new approach later. In navigating this, the PRA recognises that any change entails costs for firms, and their internal investment planning will require clarity about future PRA demands.

Additional considerations

3.24 The PRA will seek input from firms to create clearer data preparation instructions, including during the policy consultation and implementation stages. It is the PRA’s current view that a formal Q&A process, similar to that of the European Banking Authority, would not present a cost-effective approach going forward, given clearer templates and instructions, transparency in the PRA’s use cases, and effective consultation when collections are defined.

3.25 Any data collection changes would require some technology change to implement them. Nonetheless, the design of any future technology platform for data collection is beyond the scope of this DP. However, should such a need be identified, the PRA would, consistent with Principles 1 and 3 (in particular), engage on its design with firms of all types and sizes at an early stage. As well as being cost-effective and practicable, any technology solutions would need to be suitably resilient and provide ongoing data security.

3.26 As the PRA develops its approach to banking reporting, it will consider knock-on consequences for disclosure. However, disclosure is not currently within the scope of the FBD programme, as disclosure differs from reporting in its aims and market participants have different data needs from those of the PRA.

Questions

Q1: Given the PRA's objectives and data needs (as described in Chapter 1), are the principles appropriate to guide its framework for bank reporting? Are there further trade-offs that the PRA should consider in designing that framework?

Q2: How would respondents approach each of the trade-offs?

Q3: How would firms weigh an incremental vs big bang approach, given the balance of costs of change vs costs in steady state?

Developing a roadmap

3.27 The PRA expects to develop a roadmap for reforms in collaboration with industry, taking into account responses to the questions in this paper. The scoping of future phases of the FBD programme will draw on the points raised in this paper and will be subject to the PRA’s levy consultation process.

3.28 Subject to responses to this DP, areas for consideration in future phases of FBD could include:

• further rationalisation of existing reporting, extending the changes made in PS27/25 – Future banking data review: Deletion of banking reporting templates to additional whole and partial template deletions;
• further investigation of the costs of supplying data to the PRA and refinement of cost-benefit models for reporting;
• simplifications to the structure of PRA Rules on reporting; and
• refining instructions and accommodating the development of a data dictionary approach.

3.29 The PRA will also consider suggestions made in response to consultations in prioritising further work by data theme.

Questions

Q4: Do particular reporting topics stand out as starting points for review in the short to medium term, and what strategic changes would yield the biggest reduction in firm burden in the longer term?

Q5: Do firms have further information to share on which elements are the biggest drivers of firm costs (eg size of data request, frequency, complexity of instructions, validation, managing coherence, divergence of regulatory and internal definitions, lack of central dictionary)?

Q6: Do respondents have any further comments on the direction of the Future Banking Data programme?