We consider a setting in which a social planner endows an agent with a property right. The agent then makes a costly investment decision that determines the distribution of her private information. After the agent's type and the state of the world are realized a designer then designs a reallocation mechanism in a sequentially rational way, respecting the property rights held by the agent. Investment is socially efficient, but the designer cannot commit to the reallocation mechanism before the agent makes the investment decision. This provides a simple but flexible framework for analyzing the optimal design of property rights. By assigning the agent a property right the social planner reduces the severity of the agent's hold-up problem and partially restores commitment to future allocation mechanisms by guaranteeing that the agent has a sufficiently high outside option. We characterize the optimal property right as a function of the planner's and the designer's objectives, the investment technology and whether the investment is observable. We show that the optimal property right generally takes a simple form, often featuring an option-to-own and a monetary transfer to the designer.