This thesis models rational criminals and regulators with flawed incentives. In it we develop a rational model of crime and regulation that we use to show the SEC;;s current incentive structure is ineffective at preventing fraud. Under our model, criminals balance the monetary rewards of larger frauds against an increased chance of being apprehended; and regulators design regulations tominimize either the damage caused byfraud orsome other metric. We show that under this model, the SEC;;s focus on ;;stats;; and ;;quick hits;; leads to large frauds and a large social loss. We argue that regulators need to focus not just on successful prosecutions, but also on harm reduction and deterrence.