Holistic Systemic Risk Review
Trace interactions across silos before changing rules or systems
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 5
- Confidence
- 94%
The Holistic Systemic Risk Review evaluates a proposed change by widening the unit of analysis beyond the team, rule, product, or market immediately affected. First map the broader system and its participants, then examine how the change alters incentives and behavior within each part. Trace those effects across organizational, legal, technical, and economic boundaries, paying particular attention to feedback loops and risks transferred into less visible areas. The output is not merely a list of direct benefits and costs, but a connected account of how multiple factors could combine over time. This mechanism counters silo thinking, in which each participant optimizes a narrow domain while ignoring the cumulative instability produced by everyone doing the same.
Origin
Hilary Allen developed this perspective through investigating the causes of the 2008 financial crisis for the Financial Crisis Inquiry Commission. Extracted from Masters in Business.
Core principles
- 01Systemic outcomes emerge from interacting parts
- 02Local improvements can create system-wide risks
- 03Incentives and flow-on consequences matter
- 04Complex failures rarely have a single cause
- 05Broad context should precede intervention
How to run it
- 1
Define the wider system
Draw the boundary broadly enough to include adjacent markets, institutions, rules, technologies, and affected communities. Do not treat the initiating department or product as the whole system.
Pro tip Include parties that bear risk even when they do not control the decision.
Watch out A boundary drawn around only the decision-maker will conceal transferred risks.
- 2
Map components and connections
List the relevant actors, processes, dependencies, and rules, then show how resources, information, and risk flow between them.
Pro tip Ask what would stop functioning if each connection failed.
Watch out An inventory without relationships will miss the mechanisms that create systemic outcomes.
- 3
Trace changed incentives
For each affected actor, determine how the intervention changes rewards, constraints, and likely behavior. Follow the consequences into neighboring parts of the system.
Pro tip Consider how several individually rational responses could combine into a harmful collective result.
Watch out Do not assume participants will continue behaving as they did before the rules changed.
- 4
Stress-test interactions
Model how the identified effects could reinforce one another during adverse conditions. Look for concentration, feedback loops, hidden dependencies, and risks shifted outside the original silo.
Pro tip Use a crisis or failure scenario rather than testing only normal conditions.
Watch out A change that appears safe in isolation may become dangerous when several pressures occur together.
- 5
Revise from the system view
Compare the proposed change with alternatives that reduce risk across the whole system rather than merely improving one component. Record residual risks and assign responsibility for monitoring them.
Pro tip Prefer interventions that remain robust when assumptions about another silo fail.
Watch out Do not declare success solely because the initiating silo reports a benefit.
In the wild
A regulator considering lighter disclosure rules maps lenders, borrowers, securitizers, investors, and safety-net institutions. The review finds that cheaper origination could encourage weaker underwriting, transfer opaque risk into securities, and leave public institutions exposed during a downturn. The regulator retains core disclosures while simplifying only duplicative reporting.
→ Administrative costs fall without discarding protections that limit system-wide risk.
A payment company proposes removing transaction delays to improve conversion. A systemic review examines merchants, consumers, banks, fraud teams, and dispute processes and finds that the faster flow would shift losses toward small merchants. The company instead applies risk-tiered delays and monitors chargeback concentration.
→ The platform improves speed while avoiding a hidden transfer of fraud risk.
Common mistakes
Optimizing one silo
Treating a local cost reduction as a complete success ignores risks and incentives created elsewhere in the system.
Listing factors without connecting them
A broad inventory is insufficient unless the analysis explains how factors interact to produce an outcome.
Testing only normal conditions
Systemic weaknesses often become visible only when stress causes dependencies and feedback loops to activate together.
Is it for you?
Best for
It is best for regulators, leaders, and strategists evaluating changes in complex interconnected environments.
Not ideal for
It is excessive for isolated, reversible decisions with no meaningful downstream effects.
From the transcript
“there are a lot of things that come together and you need to really look very broadly to understand systemic changes.”
“part of the challenge when you're trying to prevent a financial crisis is this silo mentality where people just think about their own little piece…”
“my real takeaway was always to have the most holistic perspective possible to break down that silo mentality”
From the episode
How AI Could Freeze Progress with Hilary Allen
Hilary Allen