Trust at Risk: A Domain-Layered Model of Interstate Trust and Negative Trust Convexity

Abstract

International Relations has a developed literature on interstate trust, yet it largely treats trust

either as a relationship-level disposition — ally or rival — or as a single aggregate measure of

one state’s confidence in another. Both collapse a multi-dimensional construct into a scalar.

This paper proposes that trust between two states is better understood as a vector: a set of

distinct layers — security, trade, culture, finance, law enforcement, regulatory standards and

others — held simultaneously, each occupying a different height on one continuous spectrum

and governed by a different incentive structure. The height of any layer is set by what I term

trust at risk: the expected unrecoverable loss to a party should the counterparty defect at that

layer, operationalised as severity × irreversibility, and conceived as a deliberate transposition

of value-at-risk into the diplomatic domain. The model’s central mechanism is negative trust

convexity: the loss from a defection, and the time required to recover from it, scale non-linearly

with the height at which the breach occurs, so that the highest-trust layers carry a payoff

structure asymmetric to the trusting party’s detriment, directly analogous to negative convexity

in fixed-income instruments. The framework yields an endogenous account of contemporary

order recalibration, and because trust at risk is expressed as a priced, unrecoverable loss, it

crosses cleanly into risk and capital-allocation applications. The paper builds on the author’s

earlier work on Circles of Trust and on the dynamic weighting of international-system

influence.