This paper explores the effects of cannibalization and business-stealing within the lenders’ branch networks in the U.S. shadow banking mortgage market. Given the difficulty in obtaining data related to the shadow bank branch network, a novel dataset, the Your-Economy Time Series dataset, is introduced to provide details on shadow bank branch locations from 2012 to 2017. Implementing a two-stage least squares approach shows that having a branch increases the lender’s market share. However, the marginal effect of an additional branch on market share decreases as the total number of branches within a market increases, indicating the presence of market share cannibalization effects. In addition to market share cannibalization, there are business-stealing impacts from the presence of rival lenders’ branches. Then, using an ordered probit with a control function approach, results show that when lenders decide to expand their branch network, they consider potential cannibalization and business- stealing effects. Finally, a structural model of entry and exit is estimated to  decompose the impact of cannibalization and business-stealing effects on new branch openings. Counterfactual simulations show that muting the cannibalization and business-stealing effects leads to an increase in the establishment of new branches.