Radical Uncertainty Is a Social Problem, Not a Statistical One

Radical uncertainty is not a harder version of risk. It is a different category altogether, and the financial services industry persistently confuses the two.

Risk is what you have when you can write down the distribution. Radical uncertainty is what you have when you cannot: when the question itself is unstable, when the relevant variables have not yet been named, when the past offers analogy rather than data. Most of the decisions that actually matter to a board, whether to enter a market, how to respond to a regulator's shift in posture, whether a competitor's move is a feint or a strategy, sit firmly in the second category. And yet the apparatus around the board, the models, the scenario decks, the probability-weighted forecasts, are built almost entirely for the first.

This matters because human beings are, in fact, rather good at deciding under radical uncertainty. We have spent a long evolutionary period doing it. The capacity is real, but it does not live where finance tends to look for it. It does not live inside the individual decision-maker, refining their priors in quiet contemplation. It lives in the exchange between people: in the argument, the half-formed observation, the colleague who says the thing no one else will say, the customer who describes the world in a vocabulary the strategy team has not yet adopted. Judgement under uncertainty is a social act before it is a cognitive one.

This is the part senior leaders most often get wrong. The instinct, under pressure, is to narrow the circle. Fewer voices, faster decisions, tighter control of the narrative. It feels disciplined. It is, in fact, the precise opposite of what the situation requires. When the future cannot be modelled, the value of an additional perspective is at its highest, not its lowest. The CEO who consults three trusted lieutenants before a strategic call is not being prudent; they are starving the decision of exactly the input that would have made it better. We have sat in enough of those rooms to recognise the pattern. The decision gets made, it gets justified afterwards with the language of conviction, and the alternative readings of the situation, which existed, which were available, never reach the table.

The practical consequence is that stakeholder intelligence is not a soft input. Understanding what your major investors, your regulators, your largest clients, and your own senior people actually think, as distinct from what they say in the formal channels, is not a comfort exercise to be conducted after the strategy is set. It is a constituent part of forming the strategy in the first place. The institutions that handle uncertainty well are the ones that have built genuine mechanisms for that exchange: not surveys, not town halls, but structured access to candid views from people who have something at stake and nothing to gain from telling the leadership what it wants to hear.

The challenge for senior leaders in financial services is therefore quite specific, and quite uncomfortable. It is not to acquire better models, though better models help at the margin. It is to ask, honestly, whether the decision-making architecture around them is configured for risk or for uncertainty. If the answer is risk, and for most institutions it still is, then the next significant call, the one that will define the next three years, is being made with the wrong instruments. The fix is not technical. It is a willingness to widen the conversation before the conviction sets in, and to treat the views of those who see the institution from the outside as data of the first rank, not commentary on a decision already taken.