Loan-Level Data and the Evolution of Prudential Reporting in New Zealand

thomasverlaet
2 days ago
3 min read

Executive Summary

The Reserve Bank of New Zealand’s loan-level data (LLD) requirements represent a structural shift in prudential reporting. They move regulatory submissions away from aggregated returns toward granular, repeatable datasets that support deeper supervisory insight.

For New Zealand banks, this shift introduces new complexity. Loan-level reporting is no longer limited to static snapshots of balances and exposures. It increasingly requires alignment across credit risk, capital, accounting, and statistical reporting, as well as the ability to explain how positions evolve over time.

Successfully responding to this change requires more than technical compliance. It demands a rethink of operating models, data governance, and reporting architecture.

Why Regulators Are Moving to Granular Collections

Aggregated reporting provides high-level signals, but it inherently obscures distributional risk and behavioural dynamics within portfolios. Loan-level data allows supervisors to analyse risk at its source.

Granular collections enable regulators to:

Identify concentrations, outliers, and emerging risk patterns
Apply consistent supervisory logic across institutions
Reduce reliance on ad-hoc data requests during periods of stress
Reuse datasets across multiple analytical and supervisory purposes

This reflects a broader global trend toward data-led supervision, where the quality, structure, and traceability of data are as important as the reported outcomes themselves.

From Periodic Returns to Enduring Data Assets

Loan-level reporting shifts the focus from producing individual regulatory returns to maintaining enduring regulatory data assets.

In practice, this introduces expectations around:

Stable definitions applied consistently at the record level
Clear data lineage, from source systems through to regulatory submission
Reusability, where the same dataset supports multiple reporting obligations
Auditability, extending to individual loans rather than just aggregated totals

This change brings regulatory reporting closer to enterprise data management and risk governance disciplines, increasing the importance of structure, documentation, and control.

NZ-Specific Complexity: Aligning Risk, Capital, and Statistics at Loan Level

For New Zealand banks, loan-level data introduces additional layers of complexity that go beyond basic exposure reporting.

Alignment across ECL, RWA, and statistical reporting

Loan-level datasets increasingly need to support:

Expected Credit Loss (ECL) metrics
Risk-Weighted Assets (RWA) calculations
Statistical and supervisory data points across applications, facilities, security, customer and loan tables

While these concepts are well understood individually, aligning them at a common loan-level grain is challenging. ECL models often include management overlays or post-model adjustments that are applied at portfolio or segment level in today’s aggregated reporting. Translating these adjustments into a loan-level context requires careful design to preserve both regulatory intent and accounting integrity.

From balance snapshots to behavioural reporting

Loan-level reporting is also expanding beyond static balances. Requirements such as loan flow reporting introduce a temporal dimension — capturing what happened during the reporting period, not just where exposures ended. This includes movements such as:

Interest accrual and charges
Scheduled repayments and actual repayments
New lending and loan redraws
Reclassifications, restructures and write-offs

Supporting this type of reporting requires systems and processes that can:

Track opening and closing positions consistently
Attribute movements to defined event types
Reconcile flows back to balances in a transparent way

The result is a reporting model that increasingly resembles event-based data processing, rather than periodic extraction and aggregation.

Operational Impact: Scale, Discipline, and Sustainability

Operationally, loan-level data magnifies both strengths and weaknesses in existing reporting models.

Common challenges include:

Fragmented ownership across lending, finance, risk, and IT teams
Increased reliance on upstream data quality, with less scope for downstream correction
The need for repeatable validation and reconciliation frameworks
Higher expectations around review, sign-off, and evidencing, even at scale

Treating LLD as a one-off regulatory project often leads to fragility. Embedding it into business-as-usual processes tends to deliver more sustainable outcomes.

Technology Considerations: Designing for Evolution

The technical demands of loan-level reporting differ materially from traditional prudential reporting. They typically involve:

Larger data volumes and more complex transformations
Stronger dependencies on source systems and integration layers
Higher expectations around timeliness, resilience, and transparency

More robust approaches emphasise:

Separation between data ingestion, transformation, validation, and reporting
Configuration-driven rules to accommodate regulatory change
Strong metadata, controls, and lineage
Architectures designed for ongoing evolution, not one-off delivery

These foundations are increasingly reused across stress testing, climate risk, and cross-jurisdictional reporting initiatives.

Looking Ahead

Loan-level data should be viewed as foundational infrastructure rather than a standalone obligation. It establishes the conditions for more data-intensive supervision and more integrated regulatory reporting.

Institutions that invest early in scalable data, governance, and workflow capabilities are better positioned to:

Respond confidently to regulatory change
Reduce long-term reporting complexity and operational risk
Extract internal value from data already produced for supervisory purposes

Enabling the Shift

Supporting loan-level reporting at scale requires more than point solutions or bespoke builds. It requires platforms that can manage granular data, evolving rules, and complex workflows in a controlled and transparent way.

Reg360 is designed to support this shift — helping organisations manage loan-level regulatory data as a governed asset, while providing visibility across preparation, validation, review, and submission as requirements continue to evolve.