2006 Student Poster Presentations

small


Risk, Financing and the Optimal Number of Suppliers

Pierre-Yves Brunet

Industrial and Operations Engineering

University of Michigan

We study how supply risk, fixed supplier costs, financial constraints, and the dual role played by the suppliers as the providers of parts and the financiers of the manufacturer affect the relationship among firms in a supply chain, supplier selection, and supply chain performance. Using a one-period model, we consider joint procurement and financing decisions of a firm with limited financial capital, facing either an uncertain demand or an uncertain supply. We find that the optimal financing decisions are consistent with the financial pecking order theory and we characterize the optimal operational decisions. Our analysis suggests that the alternative financing sources (bank loans and trade credit) are complementary and that the firm uses more suppliers if the bank financing is not available. Surprisingly, we also find that the limit on the supplier loans and the wholesale price affect the optimal number of suppliers in a non-monotone way. By considering tradeoffs between the expected profit and the value of the option to default, we explain the effects of supply uncertainty on the shareholders' value and the optimal decisions. Finally, we address the question of whether the firms operating in the developing economies should contract with more suppliers than the firms operating in the developed economies. The answer is no, if the fixed cost of an extra supplier is high. However, financial constraints will force firms in the developing economies to the suboptimal level of production and cause higher stock-out rates. This is joint work with Volodymyr Babich, Goker Aydan, Pierre-Yves Brunet, Jussi Keppo, and Romesh Saigal.

small

Load Balancing Heuristic for Parallel Queues

Luz A. Caudillo-Fuentes

Industrial and Operations Engineering

University of Michigan

Many practical applications of queueing models involve heavy-tailed service times. With non-exponential services, the dynamic load balancing of a parallel processing network (PPN), a system of parallel queues, is a mathematically intractable problem. As an approximation, we consider a two station, two class exponential model with "usual" and "slow" customer types. Structural results for this model are used to develop a heuristic model for heavy-tailed service PPN's.

small

Aircraft and Crew Scheduling Problems – Sequential Approach

Jenny Diaz-Ramirez

Industrial Engineering

Tecnologico de Monterrey (ITESM)

Campus Toluca, Mexico

An approach to solve simultaneously the Aircraft and Crew Scheduling Problems is expected to get more adapted solutions for small airlines with objective values at least as good as the current available ones. First, in order to test this hypothesis, a sequential approach has been developed. Models and solution algorithms to solve them with a sequential approach are proposed.

small

Pricing Equity Default Swaps under the Jump Diffusion Constant Elasticity Volatility Model

Rafael Mendoza-Arriaga

Industrial Engineering and Management Sciences

Northwestern University

Equity Default Swaps (EDS) are financial instruments designed to bring credit protection to buyers by incorporating both: the debt issuing firm's default event and its stock price. The EDS buyer makes regular premium payments at established intervals in order to obtain protection on the event of default, being entitled to receive a recovery amount as percentage of the debt's notional at the moment the stock price reaches a lower barrier level or jumps to default. Carr and Linetsky introduced the Jump to Default Extended CEV Model (JDCEV) in order to capture the positive link between default and equity volatility factors by allowing the jump to default hazard rate to be an increasing affine function to the variance, while also capturing the inverse relationship between the volatility and the stock price. Our contribution in this work is the reduction of the JDCEV process into a standard CEV process by means of Girsanov's theorem and accurately pricing the EDS instruments by truncated α-moments of the CEV process, where α can be a negative or non integer number. The advantage of this work over recent conducted research is that while we use the JDCEV process as the underlying dynamics, the later has only been focused on the standard CEV processes and it hasn't been able to accurately formulate the Equity Default Swap instruments. Also the authors are not aware of any previous analytical solution to the truncated α-moments of the CEV process.