SUNANDO ROY – On Banking, Finance and Society

 “Getting the Balance Right”

An important shift in prudential supervision is visible in the Australian Prudential Regulation Authority’s 2026–27 Corporate Plan. APRA does not frame regulatory burden reduction as being in tension with sound supervision. Rather, it treats regulatory balance as an element of supervisory effectiveness.

The starting point is significant. APRA explicitly states that it does not pursue a “safety at all costs” agenda. Its prudential framework is, by design, aimed at minimising undue costs of regulation for industry, while its standards and supervisory practices are intended to ensure that financial institutions can fulfil their vital and productive roles in the economy.¹

This philosophy becomes more concrete under APRA’s strategic objective “Getting the balance right” — delivering financial safety and stability without undue cost for industry. APRA intends to simplify certain requirements, reduce duplication and free up capacity for lending and investment. More unusually, it states that these simplification measures should offset the impact of new requirements introduced, so that in aggregate the planned policy changes have a broadly net-neutral impact on regulatory burden.¹

This is more than a deregulatory commitment. Four elements are particularly relevant from a supervisory perspective.

First, proportionality is made observable. APRA expects larger and more complex institutions to face more stringent prudential requirements and more heightened supervisory demands. Its performance framework specifically targets Tier 1 entities being subject, on average, to significantly more formal supervisory engagements than Tier 4 entities.¹ This converts proportionality from a broad principle into something capable of being tested — and, in principle, of being failed.

Second, APRA is addressing duplication at source. For all new and substantially revised data collections, its stated target is consultation with peer agencies on data sharing. Its broader programme includes retiring, simplifying and sharing data collections where appropriate, while developing supervisory dashboards and advanced analytics to make supervision more data-driven, proportionate and timely. It is also coordinating operationally: a joint stress test with the Reserve Bank of New Zealand, and a review of small business lending sequenced with ASIC’s review of non-banks explicitly to reduce duplication for industry.¹

Third, the commitment is quantified rather than aspirational. APRA reports that harmonising bank capital rules with New Zealand will reduce annual costs for banks operating in both jurisdictions by around $175 million; it estimates long-term industry savings of approximately $6 million annually from consolidating data collections onto APRA Connect; and it intends to finalise licensing reforms that would roughly halve the time taken to process new bank licence applications.¹ Burden reduction expressed in dollars and days is burden reduction that can be audited.

Fourth, simplification does not mean lower supervisory expectations where risk is increasing. APRA simultaneously proposes deeper attention to cyber and AI risk, geopolitical risk, bank lending standards, liquidity and operational resilience — and states that it retains a strong appetite to increase supervisory intensity and take formal enforcement action where risks are material or cooperation is lacking.¹ The implicit principle is therefore important: reduce low-value supervisory friction so that supervisory capacity — and licensee management attention — can move toward material risks.

That distinction is critical. The appropriate objective is not less supervision. It is less unnecessary supervision.

Emerging Global Practice

APRA’s approach is  consistent with a wider international movement. Following years in which financial regulation accumulated layer upon layer of rules, returns, thematic exercises, questionnaires and ad hoc requests, regulators are beginning to examine the stock as well as the flow of supervisory requirements.

The European Banking Authority provides perhaps the clearest current example. On 10 April 2026 it launched what it describes as an unprecedented simplification package for EU supervisory reporting. The proposals would reduce the number of data points across EU harmonised reporting by around 50 per cent — a net reduction achieved despite the simultaneous addition of new requirements for IFRS 18, ESG and the Fundamental Review of the Trading Book. Separate EU-wide stress test and supervisory benchmarking collections would be folded into regular reporting, and proportionality strengthened for small and non-complex institutions.²

The scale of the underlying problem is revealing. The EBA’s stocktake of national data collections found that in 2025, EEA competent authorities made 671 structured microprudential data requests to banks and their branches, totalling 978,109 data points — concentrated, notably, in areas already well covered by the harmonised reporting framework, with credit and counterparty risk alone accounting for 226 requests and some 709,865 data points. Competent authorities have already implemented or planned simplifications reducing national reporting data points by 17.4 per cent, and 18.5 per cent for SSM data requests. An EU-wide public repository of supervisory data requests is expected to be operational in early 2027.³

At the ECB, the Governing Council endorsed in December 2025 the recommendations of its High-Level Task Force on Simplification, which include increasing the risk focus of supervision by changing the level of prescriptiveness of regulation governing supervisory processes, and — in the reporting domain — a “request once” principle of stepped-up coordination and data sharing among authorities, alongside a “report once” integrated reporting system. The framing is deliberately cautious: resilience is to be maintained, and simplification is not deregulation.⁴ ECB Banking Supervision has since described its own agenda in the same terms — streamlining supervision while safeguarding resilience.⁵

The United Kingdom illustrates another direction. The PRA’s Strong and Simple framework tailors capital, liquidity and reporting requirements for small domestic deposit takers while seeking to preserve overall resilience, with the simplified capital regime and more proportionate reporting taking effect from 1 January 2027. Its 2025–26 reporting on competitiveness and growth introduced three new metrics, including an explicit one on the administrative burden on firms, and retired three metrics that duplicated information published elsewhere or yielded no meaningful insight — deletion applied to the supervisor’s own performance framework as well as to firms’ returns.⁶ Separately, the Future Banking Data programme deleted 37 banking reporting templates with effect from 31 December 2025, targeting templates that were duplicative, outdated or of limited supervisory value.⁷ The PRA has also moved all firms to a two-year Periodic Summary Meeting cycle — a reduction in the frequency of engagement rather than its content.⁶

Hong Kong has similarly pursued legislative amendments intended both to improve regulatory clarity and effectiveness and to reduce compliance burden; the Banking Legislation (Miscellaneous Amendments) Bill was gazetted in June 2026 following consultation conclusions published in February.⁸

These developments suggest an emerging global model:

This direction is consistent with the Basel framework itself. The Core Principles for Effective Banking Supervision, as revised in 2024, embed proportionality more explicitly than their predecessors, while the accompanying Basel and FSI literature is clear that proportionality is about calibrating supervisory intensity to risk and systemic importance — not about diluting prudential safeguards for smaller or simpler institutions.⁹

The Road Ahead: A Supervisory Stance

The next stage should move beyond periodic regulatory simplification exercises. Licensee burden should become a normal consideration in supervisory design and decision-making.

This does not mean that a supervisor should hesitate to demand information when risk warrants it. A weak bank, rapidly growing institution, material control failure or emerging systemic vulnerability may legitimately face very substantial supervisory demands. The key question is whether each demand has a clear supervisory purpose proportionate to the risk being addressed.

A useful future supervisory stance could therefore rest on seven disciplines:

1. Risk determines intensity. Supervisory demands should increase demonstrably with risk, complexity, systemic importance and unresolved weaknesses — not simply because an institution is available to supervise.
2. Request once, use many times. Before seeking information, supervisors should establish whether it already exists in regulatory returns, previous inspections, prudential databases or another authority. The EBA’s finding that the bulk of national ad hoc requests fell in areas already covered by harmonised reporting is the cautionary case.
3. Apply a materiality test to supervisory requests. A request should reasonably be expected to affect risk assessment, supervisory judgement or action. Information gathered merely because it might be useful has a cumulative cost.
4. Recognise change burden. Ten individually reasonable initiatives arriving simultaneously can collectively become unreasonable. Supervisory planning should therefore consider the aggregate portfolio of inspections, thematic reviews, surveys, remediation programmes and regulatory changes confronting an institution.
5. Create a burden budget. APRA’s commitment to broadly offsetting new requirements with simplification offers a useful conceptual model, and it is no longer unique: the PRA is targeting a net reduction in administrative burdens of around £100 million, having already removed roughly £26 million a year of banks’ reporting costs through the first phase of Future Banking Data. Supervisors could periodically ask: what can be stopped, shortened, automated or reused when a new demand is introduced?
6. Measure both sides of proportionality. Supervisors already measure risk, remediation and compliance. They should increasingly track duplicated requests, response hours, data points collected, frequency of engagement and the proportion of information actually used in supervisory decisions. The instruments now exist — APRA’s Tier 1 versus Tier 4 engagement metric, the PRA’s administrative burden metric, the EBA’s data-point count — and what is measured tends to be managed.
7. Do not compromise on material risk. Burden reduction should end where prudential significance begins. Poor governance, deteriorating capital or liquidity, misconduct, cyber vulnerability or persistent remediation failure should justify heightened scrutiny regardless of compliance cost.

The mature supervisory question is therefore no longer “Are we asking enough?” It is “Are we asking for the right things, from the right institutions, at the right frequency — and are we using what we ask for?”

The objective should therefore be maximum supervisory insight per unit of regulatory burden. That is not lighter supervision. It is more disciplined, proportionate and ultimately more effective supervision.