Foreign Exchange Intervention: Identification and Evidence

Foreign exchange intervention is widely used across low, middle and high income economies, yet its effectiveness remains disputed. The central difficulty is identification: interventions are typically endogenous to exchange rate movements, often bundled with other policies and sometimes unobserved. This survey reviews five decades of empirical research and reframes the debate as a problem of causal identification rather than a binary question of whether intervention works. The literature is organized by identification strategy: event studies, reaction-function residuals, controlled regressions, instrumental variables, high-frequency approaches, SVARs, structural portfolio-balance models and natural experiments. For each design, the source of identifying variation and the key assumptions required for causal interpretation are specified. Differences in reported effectiveness largely reflect differences in design, data and regime context. The overall finding is that intervention does affect exchange rates: evidence is strongest for short-run effects on returns and volatility, especially when operations are large, well-timed and implemented in thinner markets. Persistent exchange rate level effects are more sensitive to assumptions about asset substitutability, expectations and financial frictions. The survey provides guidance for empirical practice and highlights priorities for improved measurement and future research.