How do external loss data consortia like ORX work?

Question

AcadiFi · Accepted Answer

The Operational Riskdata eXchange (ORX) is the largest global banking consortium for pooling operational loss data, with over 100 member banks sharing anonymized loss events...

Unlock with Scholar — $19/month

Get full access to all Q&A answers, practice question explanations, and progress tracking.

Start 7-Day Free Trial Compare Plans

No credit card required for free trial

How it works: member banks submit loss events quarterly following standardized classification rules, with events above approximately $25,000 required. ORX anonymizes and distributes the pooled database back to members. Each event record includes event type, business line, gross loss amount, occurrence and discovery dates, and descriptive narrative scrubbed of identifying details. Members receive data covering roughly 800,000+ events accumulated over two decades.

Usage in capital modeling: banks use external data to calibrate severity distribution tails. Since a mid-size regional bank may never have experienced a $500 million rogue-trader loss, its internal data alone would underestimate such risk. External data fills this gap. Standard techniques include scaling external losses by institutional size (using asset ratios or revenue ratios), filtering for relevance by business line and geography, and splicing external-data tails onto internal-data bodies through techniques like extreme value theory (EVT).

Scaling challenges: a $200 million loss at a global universal bank does not translate to a $200 million loss at a community bank. ORX recommends size-scaling adjustments such as dividing gross loss by gross income and multiplying by the receiving bank's gross income, though research shows scaling relationships vary by event type.

Other consortia: Algo OpData and regional consortia complement ORX. US national data collection through the ABA Banking Loss Database provides similar pooling. For FRM exam purposes, remember that the AMA's four mandatory inputs include internal loss data, external loss data, scenario analysis, and BEICFs. External data's unique role is tail calibration — events too rare for any single bank's history but observable in the industry at large. Without it, OpVaR estimates would systematically understate tail risk.

How do external loss data consortia like ORX work?

Master Part II with our FRM Course

Related Questions

Practice Questions