Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

WD23

2 - Life Cycle Pricing of Multiple Differentiated Products Hongmin Li, Arizona State University, WP Carey School of Business, Dept of Supply Chain Management, Tempe, AZ, 85287, United States This paper considers the centralized pricing problem of a firm managing multiple substitute or differentiated products. Demands of these products undergo a diffusion process and customers choose among the products, with the choice probability of each product given by the logit model. We examine the firm’s optimal pricing problem in this context. 3 - Dynamic Pricing and Replenishment for Seasonal and Regular Products Oben Ceryan, LeBow College of Business, Drexel University, Philadelphia, PA, 19103, United States We consider a firm that offers two substitutable products that differ in how their inventories are managed, a seasonal product with a fixed initial quantity that allows dynamic price adjustments but no replenishments, and a regular product with a static price that can be periodically replenished. We study the impact of these asymmetries on optimal dynamic pricing and replenishment decisions and utilize the insights gained to develop a simple-to-implement and effective heuristic policy. 4 - Competitive Revenue Management with Sequential Bargaining We study the role of bargaining in competitive revenue management. The seller firms each have some fixed initial stocks and compete to sell to a random arrive stream of potential buyers who are heterogeneous in product valuation. We fully characterize the equilibrium outcome, which depends on the length of the selling season and the initial stock levels. When the selling season is up to three periods, the seller with a lower stock level is weakly preferred by an incoming buyer. When the selling season is longer than three periods, the seller with a higher stock level may win the competition. n WD21 North Bldg 129B Pricing and Revenue Management Contributed Session Chair: Kimitoshi Sato, Kanagawa University, Yokohama, Kanagawa, Japan 1 - Managing Opportunistic Consumer in Retail Operations Tolga Aydinliyim, Associate Professor, Baruch College, One Bernard Baruch Way, Department of Management, New York, NY, 10010-5585, United States, Mehmet Sekip Altug, Aditya Jain Considering “honest-customers versus “renters, we assess the merits “targeted- refunds and “menu-of-refunds proposals retailers use to mitigate opportunistic consumer returns. Relative to benchmarks with “no-renters (Su, 2009) and “uniform-refunds, we find that the menu-of-refunds proposal with a low restocking-fee can separate customer types when leftover/returned units must be significantly marked down. 2 - The Effect of Internet Bots on Dynamic Pricing of Perishable Products Kimitoshi Sato, Kanagawa University, Yokohama, Kanagawa, Japan We consider the problem of a firm that sells perishable products with a dynamic pricing scheme in the presence of internet bots. The bots automatically check for changes in price every few seconds and hold the price during short period (e.g. 24 hours). Reservations by the bots does not generate revenue for the firm if payments are not completed and will temporarily increase the selling price. In this research, we formulate a dynamic pricing model in presence of the bots and derive an optimal pricing policy. Then we show how does the behavior of bots affect the firm’s expected revenue as well as the optimal price trend. n WD22 North Bldg 130 Practice- E-Business/Commerce I Contributed Session Chair: Roger William Baugher, TrAnalytics, LLC, Johns Creek, GA, 30097, United States 1 - The Moderating Effect of Reputation and Presentation on Trust Transfer of Trusted Third Party Cong Cao, University of Wollongong, Wollongong, NSW 2522, Australia, Jun Yan, Mengxiang Li Yuanchen Li, Purdue University, 2192 Tortuga Lane, West Lafayette, IN, 47906, United States, Qi Feng, J. George Shanthikumar

Based on experimental data, this paper not only shows the significant influence of trusted third party (TTP) on consumers’ behavior, but further explains its trust transfer function in the online shopping environment. Moreover, it also reveals moderator variables of the trust transfer, namely the reputation and presentation of TTP. The research results show that the certification service provided by TTP with high reputation can significantly improve the degree of consumer trust in enterprises. The concrete and detailed description and presentation of TTP services can effectively enhance consumers’ cognition of TTP services and therefore influence their behavioral intention. 2 - End-to-end Inventory Replenishment Model Yuanyuan Shi, University of Washington, 6319 65th Avenue NE, Seattle, WA, 98115, United States, Di Wu, Meng Qi, Rong Yuan, Yuhui Shi, Max Shen Traditional inventory models often assume a predict-then-optimize paradigm. A prediction model which depicts the demand uncertainty is built then an optimization model is applied to solve for optimal inventory decisions based on the forecasting. However, the criteria by which we train the prediction model often differs from the ultimate criteria on which we evaluate them. In this research, we focus on building an end-to-end inventory replenishment model which simultaneously learns the probabilistic info of multiple uncertainty sources such as demand, lead time, while making replenishment decisions that directly capture the cost-based objective. 3 - Competitive Intelligence Analysis from Customer Online Concerns of Series Products for Engineering Design Jian Jin, Beijing Normal University, No. 19, Xinjiekouwai street, Beijing, 100875, China, Qian Geng, Ping Ji Online concerns inform valuable messages to customers, dealers and product designers. Conventional studies mainly focus on the sentiment analysis of online opinions and few explore their capacities on comparison of series products. In this study, a framework is presented to highlight shared pros and cons of series products by mining online concerns. Customer opinions of specific features across series products are initially extracted and clustered to identify similar concerns. Then, an optimization problem is formulated to sample a few representative sentences for designers. This study aims to integrate big consumer data for competitive intelligence analysis in market driven product design. 4 - Current State of Railroad Yard Technology Roger William Baugher, President, TrAnalytics, LLC, Johns Creek, GA, 30097, United States Given the high cost of improvements to a rail yard’s infrastructure, railroads are investigating and implementing new control technology and enhanced information systems to improve performance and reduce costs. These efforts will be discussed and a summary presented. n WD23 North Bldg 131A Quantitative Methods in Financial Engineering Sponsored: Finance Sponsored Session Chair: Qi Wu, The Chinese University of Hong Kong, Hong Kong Co-Chair: Lingfei Li, Chinese University of Hong Kong, Shatin, N. T, Hong Kong 1 - A Multidimensional Hilbert Transform Approach for Barrier Option Pricing and Survival Probability Calculation Jie Chen, Chinese University of Hong Kong, Shatin, N. T, Hong Kong, Liaoyuan Fan, Lingfei Li We propose a novel approach for pricing discretely monitored multi-asset barrier options and computing joint survival probability in multivariate exponential Levy asset price models using multidimensional Hilbert transform. Our method features exponential decay in the dicretization parameter and various numerical examples are provided to demonstrate its efficiency in both 2D and 3D cases. 2 - Parsimonious Learning of Tail Dynamics Qi Wu, City University of Hong Kong, Kowloon, Hong Kong, Xing Yan We propose a semi-parametric machine learning approach to study the serial dependence of conditional distribution in asset returns, motivated by the problem of tail risk forecast. This approach combines the merits of a novel parameterization of conditional quantile function with the Long short-term memory (LSTM) sequential neural network. Comprehensive empirical study shows that the dynamics of return distribution is not always auto-regressive in nature; further, drivers of higher moments’ serial dependencies could contain risk factors that are independent of those responsible for volatility clustering. These findings should have important implications for financial risk management.

501