FAA-Style Financial Product Safety Review
Test complex financial products before scale creates systemic risk
- Difficulty
- Expert
- Time to result
- ~ongoing to results
- Steps
- 7
- Confidence
- 93%
This model borrows the safety logic of aviation regulation for complex finance. A new aircraft is tested before broad operation, traffic is constrained to avoid dangerous concentration, and international coordination governs flights that cross borders. Applied to finance, a new complex product should undergo independent testing of its assumptions, behavior under stress, hedging dynamics, and possible feedback effects before widespread distribution. Oversight should examine not only whether one institution can hold the product, but also how many institutions can accumulate the same exposure without creating systemic concentration. Deployment limits and monitoring should evolve as usage grows. Because financial products and counterparties cross national boundaries, the review must include coordinated international standards rather than fragmented national supervision.
Origin
Extracted from Masters in Business, where Paul Wilmott described an airline-regulation analogy presented in The Money Formula.
Core principles
- 01Complex financial products deserve pre-deployment safety scrutiny
- 02A product's individual risk differs from its system-wide concentration risk
- 03Testing must include failure behavior, not just normal operation
- 04Cross-border markets require cross-border oversight
- 05Regulation should constrain dangerous scale before a crisis
How to run it
- 1
Specify the product
Document the product's economic purpose, payoff structure, counterparties, dependencies, and intended users.
Pro tip Require a plain-language explanation alongside mathematical documentation.
Watch out Complexity can obscure whether the product serves a real economic need.
- 2
Test normal operation
Evaluate pricing, liquidity, settlement, hedging, and counterparty behavior under expected conditions.
Pro tip Use independent reviewers rather than relying solely on the product designer.
Watch out Passing normal-condition tests says little about crisis behavior.
- 3
Stress the failure modes
Simulate extreme moves, correlation shifts, liquidity loss, model error, and crowded hedging responses.
Pro tip Include scenarios where many holders act on the same model simultaneously.
Watch out Static stress tests can miss positive feedback created by dynamic hedging.
- 4
Measure concentration
Aggregate exposures across institutions, counterparties, and products to identify common dependencies and crowded positions.
Pro tip Track economically equivalent exposures, not just identical product labels.
Watch out Institution-by-institution compliance can coexist with system-wide danger.
- 5
Set deployment limits
Restrict product scale, eligible users, leverage, or concentration until evidence supports broader use.
Pro tip Tie limits to transparent risk indicators and review dates.
Watch out Approval without enforceable limits may permit risk to compound unnoticed.
- 6
Coordinate internationally
Align standards and information sharing across jurisdictions touched by the product and its counterparties.
Pro tip Define responsibility before a cross-border failure occurs.
Watch out Regulatory gaps invite risk to migrate to the least restrictive jurisdiction.
- 7
Monitor and recertify
Reassess the product as market structure, ownership concentration, and observed behavior change.
Pro tip Treat rapid adoption as a reason for more scrutiny, not proof of safety.
Watch out A product that was manageable at small scale may become dangerous when crowded.
In the wild
Before allowing broad sale of a new leveraged credit derivative, reviewers map counterparties, test default clusters, simulate disappearing liquidity, and model simultaneous hedging by major dealers. They discover that many institutions would rely on the same protection sellers and impose concentration limits plus enhanced reporting.
→ The product can be introduced gradually without allowing a hidden common exposure to become systemically dominant.
Dynamic portfolio insurance appeared to protect individual portfolios, but widespread selling into a falling market created positive feedback. An FAA-style review would examine the aggregate behavior of many users following the same rule rather than testing one portfolio in isolation.
→ System-level simulation exposes a risk that an institution-level model overlooks.
Common mistakes
Testing firms in isolation
A product may appear manageable for each holder while becoming destabilizing when many holders respond identically.
Certifying only the mathematics
A mathematically coherent model can still fail because of liquidity, psychology, incentives, or market feedback.
Treating approval as permanent
Changing scale, concentration, and market structure can invalidate an earlier safety judgment.
Is it for you?
Best for
It is best for regulators, exchanges, clearing organizations, and financial institutions approving complex products at scale.
Not ideal for
It is not designed for simple instruments whose risks are already transparent, limited, and well understood.
From the transcript
“whenever you have a new plane, it has to be tested, etcetera, etcetera. Well, if you have a new financial product, it has it should…”
“you can't have too much concentration risk, just like you can't have too many planes landing the same up at the same time.”
“Planes take off one country, land another country, well, how hard can it be to have an international perspective to the banking regulation?”
From the episode
Paul Wilmott Has Some Feelings About Quantitative Models
Paul Wilmott