Government interventions such as bailouts are often implemented in times of high uncertainty. Policymakers may therefore rely on information from financial markets to guide their decisions. We study a model in which a policymaker learns from market activity and traders have high private stakes in the intervention. We discuss how the presence of such traders affects intervention outcomes, and show that it reduces market informativeness and the efficiency of bailouts. Regarding normative implications, we show that a higher social cost of interventions and a gradual implementation of assistance can improve market informativeness and raise overall welfare.