Chen, Yibin2016-12-192016-12-192016-10https://hdl.handle.net/11299/183390University of Minnesota Ph.D. dissertation. October 2016. Major: Industrial and Systems Engineering. Advisor: Diwakar Gupta. 1 computer file (PDF); ix, 137 pages.This thesis contains essays addressing three different problems in supply chain finance, studying two important source of financing – trade credit and crowdfunding. In the first essay, we model a web-based retailer offering direct financing to its capital-constrained supplier who sells its products on a consignment basis. The supplier has access to bank financing and the retailer can specify debt-seniority. We show that if the retailer requires debt-seniority, then the supplier will use either bank or direct financing (but not both), and it may produce more than the first-best quantity. In contrast, if the retailer does not request seniority, the supplier may take both bank and direct financing, but it always produces less than the first-best quantity. We find that the retailer does not always prefer debt-seniority. In the second essay, we study variants of net-terms trade credit contracts in a traditional market, with market frictions of bankruptcy costs, information asymmetry and transaction costs. We provide a possible explanation for the prevalence of debt forgiveness in practice. We show that the supplier, by contracting and preannouncing forgiveness, can use one type of friction to mitigate another, and benefit both supply chain partners. The third essay studies the social proof effect in the rapidly diffused crowdfunding market. We study how prior capital accumulation and the volume of owner-issued referrals impact subsequent fundraising. By analyzing a campaign-level web traffic data set, we show that higher capital accumulation increases the probability of contribution for the owner-referrals, the volume of organic visitors and the total volume of organic contributors, although it may decrease probability of contribution from the organic visitors. To study the optimal policy for entrepreneurs to use their referrals, we build a Markov-decision-process model and calibrate its parameters using our data. Our finding indicates that entrepreneurs in general should distribute their referrals over the course of the entire campaign.enContractsCrowdfundingSupply ChainTrade CreditThesis or Dissertation