Regional patterns in PFMI disclosures affect risk assessment

📰 Welcome to the Newsletter

Welcome to Global Custody Pro, read by custody professionals like you. I'm Brennan McDonald, Managing Editor. I write about the global custody industry, having spent over 12 years in financial services, including working at a global custody bank. An AI voice reads the audio version of this newsletter. Have feedback? Just reply to this email or connect with us on LinkedIn.

Table of Contents

• Building a PFMI Database: Early Findings

• The Data Structure

• Practical Implications

Do you enjoy Global Custody Pro? Share it with a colleague today:

About This Monday Series

Premium subscribers to Global Custody Pro receive exclusive Monday observations from this developing database project. While Wednesday and Friday bring news about custody and digital assets, Monday focuses on the post-trade infrastructure underlying global markets.

The database currently covers 23 major CCPs with quarterly data through Q2 2025. As I add more institutions and quarterly updates, subscribers get early access to observations and questions emerging from the data.

Premium subscription includes Monday analysis plus full access to all Global Custody Pro content.

Building a PFMI Database: Early Findings

I’ve been building a comprehensive database of PFMI quantitative disclosures. The same risk metric can have up to 42 different descriptions across clearinghouses. Not variations of different metrics, but 42 descriptions for the identical standardised field.

The data shows text in roughly 40% of numeric field values. Coverage ratios appear as "Cover 2", time periods as "00:00:00", percentages as "100.00%" strings, amounts followed by methodology notes. Approximately half of all disclosures include written explanations alongside numerical data. 82% of institutions reference external documentation.

Geographic patterns are remarkably consistent within jurisdictions. Four major clearinghouses in Hong Kong use nearly identical field structures. Similar alignment appears within Australia, Singapore, and other markets. Cross-border is where interpretations diverge. The framework is structurally consistent but semantically varied.

The Data Structure

Processing these disclosures revealed three elements: standardised quantitative fields (60% of values), written explanations embedded throughout (40% of values), and references to external technical documentation (in 82% of institutions). This isn't poor data quality. It's institutions providing context that standardised templates can't capture.

The written content is substantial. Some institutions provide over a hundred explanatory comments per quarter. Others embed methodology notes. Many link to external technical documents. These aren't generic disclaimers but specific explanations of calculations and assumptions.

Regional consistency suggests coordinated local implementation. Different regions adapted global standards to existing frameworks, creating distinct but internally consistent approaches. Within regions, clearinghouses basically speak the same language. Between regions, translation is needed.

Practical Implications

Comparing metrics across clearinghouses requires understanding these variations. The same number at different institutions may represent fundamentally different calculations: peak versus average, with or without haircuts, different time horizons. Without processing both numbers and written context, comparisons may mislead.

Institutions providing extensive methodology explanations often report more conservative numbers. Those with minimal explanations might use different assumptions. The numbers alone don't reveal these differences. The clearinghouse that looks riskiest by the numbers might be documenting conservative calculations, whilst one with better metrics might be using optimistic assumptions.

Upgrade to a Premium Subscription to read the rest.