Risk is not random. It has structure, and that structure determines which interventions reduce it and which merely move it.
Risk as Structure
The intuitive model of risk treats it as a property of individual events — the probability that a specific bad outcome will occur, multiplied by its magnitude if it does. This model is adequate for evaluating isolated risks in simple environments. It is inadequate for the class of risks that most institutions actually face, which are not isolated events but structured phenomena — risks that arise from the interaction of multiple factors, that concentrate in specific parts of institutional systems, and that respond to interventions in ways that the event-probability model does not predict.
The structure of risk in an institutional context has three primary dimensions. The first is concentration: where, within the institution's operations, risks are aggregated rather than distributed. The second is correlation: which risks tend to materialise together rather than independently, so that the occurrence of one increases the probability of others. The third is propagation: how risks, once they materialise, travel through the institution's systems — which failure modes trigger cascades rather than remaining contained.
Why Structure Matters for Intervention
Understanding risk structure matters because interventions that address the symptoms of risk without addressing its structure do not reduce institutional risk — they relocate it. The institution that reduces operational risk in its primary business by transferring it to a supply chain partner has not reduced its total risk exposure if that partner's failure would produce the same operational impact as the original risk would have. The institution that hedges financial risk through derivative instruments has not reduced its financial risk if the counterparties to those instruments are themselves exposed to the same underlying risk that the hedges were designed to protect against.
Risk interventions that address structure — that reduce concentration, reduce correlation, or interrupt propagation pathways — genuinely reduce institutional risk exposure. Risk interventions that address only the visible expression of risk without addressing its underlying structure produce the appearance of risk management while maintaining or sometimes increasing the underlying exposure.
Building Risk Literacy
Risk literacy — the capacity to read risk structure rather than simply estimate event probabilities — requires investment in analytical frameworks that most professional development programmes do not provide. It requires understanding systems dynamics: how systems with feedback loops can amplify rather than dampen shocks. It requires understanding correlation structures: which risks are genuinely independent and which appear independent but share underlying drivers. And it requires the discipline to look beyond the most recently visible risk — the one that produced the last crisis — to the structural conditions that determine which risks will next become visible.
Risk managed at the level of events produces a game of whack-a-mole. Risk managed at the level of structure addresses the conditions that produce the events — which is a different game with a different and more achievable objective.
◆ Continue the Argument
Risk maps to transition. Here is the full picture.
The Risk Architecture series maps the structure of individual and institutional risk. The Transition State Arc maps what happens when those risks materialize at scale — how institutions weaken, misdiagnose, and reorganize under pressure. 12 essays. Read in sequence.
Read the Transition State Arc →
Discussion