We show that exposure from past business transactions – risk overhang – can reduce activity in related business lines, sometimes to the point where no new trade occurs. We focus primarily on the role of overhang in nonlife insurance market disruptions. Our model predicts that the relative impact, duration, and character of supply disruptions are related to the extent of overhang. These predictions are consistent with differences between the mid-1980s liability insurance crisis and the early-1990s catastrophe reinsurance crisis. The overhang model predicts attributes of these crises that prior explanations relying on adverse selection do not. We also discuss applications of the overhang model to disruptions in lending and securities markets.