FRM Handbook Ch 27: Firmwide Risk Management (Integrated Risk Management)
Book: Financial Risk Manager Handbook Plus Test Bank
Author: Philippe Jorion
ISBN: 978-0-470-90401-5
By Chapter 27, Jorion stops talking about individual risk types and starts talking about how they connect inside a real institution. Firmwide risk management sounds like a buzzword until you read about UBS losing $19 billion on subprime-related securities scattered across four different books with no firmwide concentration monitoring. That is what siloed risk management looks like in practice.
Why integration matters now
Three forces push banks toward firmwide views:
- Global expansion exposes institutions to more correlated risk sources.
- Products link market, credit, and operational risks in ways that do not fit neat categories.
- Regulators and investors want a single picture of total risk, not three separate reports that nobody adds up correctly.
If you measure market risk tightly but credit risk loosely, trades migrate to wherever capital charges are lowest. Risk flows to the weakest measurement. That is not a theory. It is what happened with subprime CDO warehousing and with liquidity risk before 2008.
Risk categories and their overlaps
Jorion lists the standard buckets: market, credit, operational, business, and liquidity risk. But the boundaries blur constantly.
Collateral payments in swaps reduce credit risk but increase operational and liquidity demands. A data entry error on swap terms creates wrong market risk numbers and wrong credit exposures at the same time. A rogue trader is operational risk that becomes market risk the moment unauthorized positions hit the book.
Wrong-way trades are a sharp example of market-credit interaction. If a bank loses money on a swap with a speculator, credit risk is low. If the bank makes a large profit, the speculator may be losing enough to default. The trade that looks safest from a credit perspective is sometimes the most dangerous.
Aggregating risk across silos
Once each risk type is measured, the hard part is combining them. Jorion walks through the hierarchy:
- Convert all measures to the same currency and horizon (usually one year).
- Summarize each distribution with a VaR-type metric at a common confidence level.
- Combine across risk types.
Most banks take the simple route: add market VaR + credit VaR + operational VaR. JPMorgan’s example sums to $74 billion in economic capital. The problem is this assumes the worst loss hits all three categories at once. That almost never happens. Simple summation overstates total risk.
Alternatives include:
- Fixed diversification percentage: Subtract a flat discount (say 20%) from the sum.
- Variance-covariance weighting: Use estimated correlations between risk types. Banks typically assume high correlation between market and credit, lower for business risk, very low for operational.
- Copulas and full simulation: More flexible, much harder to implement. Few banks go here.
Economic capital and regulatory capital
Economic capital is the loss a firm is willing to absorb at a given confidence level over a given horizon. It is a VaR concept applied firmwide. Regulatory capital is what supervisors require. The two should relate, but they are not identical.
Economic capital drives internal pricing, limits, and performance measurement. Regulatory capital drives compliance. When they diverge too far, you get distortions in business decisions.
The UBS lesson
The subprime losses at UBS appeared in the CDO warehousing book, the trading book, the Treasury book, and a hedge fund subsidiary. No one tracked gross or net firmwide concentration in subprime-backed senior tranches. Each desk’s risk system showed low numbers because the models treated AAA tranches as nearly risk-free.
Firmwide integration is not about building one giant model. It is about seeing the whole exposure map before the losses arrive.
What stuck with me
Adding up three VaR numbers is not firmwide risk management. It is arithmetic with a false sense of precision.
The chapter’s real value is showing where risk types leak into each other and why banks that measure one type well but ignore another end up surprised. Integration is messy, expensive, and politically hard. The alternative is finding out about your $19 billion problem on the front page.
Previous: FRM Handbook Ch 26: Liquidity Risk
Next: FRM Handbook Ch 27: Firmwide Risk Management (Organization and RAROC)