Informs Annual Meeting 2017

WA61

INFORMS Houston – 2017

2 - Case Study: Challenges in Assembly Line Design Balaraman Rajan, Assistant Professor, California State University East Bay, 25800 Carlos Bee Blvd, Hayward, CA, 94542, United States, balaraman.rajan@csueastbay.edu, Ravichandran Narasimhan In this case study we discuss the production plan of two different assembly lines and how the activities can be reorganized to meet an increase in demand. The case can be used for teaching topics in operations and supply chain management at both undergraduate and graduate level. The first part of the case focuses on topics including capacity and bottleneck analysis, line balancing, and capacity requirement analysis. The second part of the case involves Just In Time manufacturing and developing a production plan. The case also demonstrates the usefulness of process flow and precedence diagrams. 3 - A Simulation Companion for ‘the Goal” Match/Dice Game Tom Brady, Purdue University, 1401 S US.Hwy 421, Westville, IN, 46391, United States, tbradyjr@pnw.edu Since its release in 1986, The GOAL has become the seminal book in Operations Management. One of the most oft-cited passages of the book is the Boy Scout camping trip where the main concept of the book, statistical fluctuations and dependent events are introduced through a simple match/dice game. While simple in concept, most people who read the book do not get the full understanding of the concepts. I have used a physical and computer-based simulation of this game at the undergraduate, graduate, and executive levels and have seen the level of understanding greatly improved. The simulation has allowed many embellishments to be added to the simple game that extend the learning concepts in many directions. 4 - A Best Practice Approach for a Hybrid Delivery of an Undergraduate Operations Management Course Dorothy Klotz, Professor of Management Systems, Fordham Univeristy, 140 West 62nd Street, New York City, NY, 10023, United States, klotz@fordham.edu The modular design of both the course and the learning materials facilitated the successful transition of quantitative content online while reserving classroom time for active learning exercises. Online technologies were used to pace student learning and maintain student accountability. This approach not only resulted in a dramatic reduction in student seat time, but also resulted in improved learning outcomes. 370B Finance, Portfolio Analysis Contributed Session Chair: Xiaoling Mei, Xiamen University, Xiamen, Fujian, China, meilingluck@gmail.com 1 - Bayesian Dynamic Linear Models for Strategic Asset Allocation Jared D.Fisher, University of Texas, 2110 Speedway Stop B6500, Austin, TX, 78712, United States, jared.fisher@phd.mccombs.utexas.edu, Carlos Carvalho, Davide Pettenuzzo We find statistical and economic evidence of the predictability of risk premia for multiple risky assets. We present a unifying model that accounts for parameter uncertainty, stochastic volatility, and time-varying parameters through a Bayesian dynamic linear model framework with variance discounting. To account for model uncertainty, we average over many potential models of different combinations of predictors and discount factors. For an investor allocating his wealth between a risk-free asset, stocks, and bonds with maturities from 2 to 10 years, we demonstrate an ensemble of features that is preferable to others, including the models assumed by the efficient markets hypothesis. 2 - The Curious Case of Negative Volatility Christoph Merkle, Associate Professor, Kuehne Logistics University, Grosser Grasbrook 17, Hamburg, 20457, Germany, christoph.merkle@the-klu.org Investors mostly perceive their own portfolio as no more volatile than the market portfolio. Taking into account observed portfolio betas, this implies a belief in very low idiosyncratic portfolio volatility, which is even negative for a considerable fraction of the population. Possible explanations are extreme overconfidence in combination with a misunderstanding how market and portfolio volatility are related. The identified bias strongly contributes to underdiversification, as a belief in negative idiosyncratic volatility conceals the benefit of diversification. WA61

sell-securities strategy.We extend the results to optimally allocate money in pairs of securities and riskless asset and transform the primal optimization problem into a quadratic optimization problem with a closed-form solution. 4 - Using Social Media Sentiment in Automatic Portfolio Rebalancing Jing-Rung Yu, National Chi-Nan University, Nantou, Taiwan, jingrungyu@gmail.com, Wan-Jiun Paul Chiou, Cing-Hung Hung, Wen-Kuei Dong, Yi-Hsuan Chang Studies have documented the impact of the social media on stock pricing. We analyze sentiments by applying the Stock Market Lexicon to quantify the posted opinions in the Twitter combining with various portfolio models to construct optimal asset allocation. Using the daily returns and the tweets of the component stocks in S&P 500 Index during July 1, 2016 to January 31, 2017, we rebalance portfolio by identifying the 10 most tweeted stocks in previous three days. The empirical results show the dynamic rebalancing portfolios incorporating social media opinions outperform the fixed rebalancing portfolios in various measures. 370C Reliability Contributed Session Chair: Hung Do, University of Vermont, Burlington, VT, United States, hdo@uvm.edu 1 - Performance Analysis of Finite Production Run-based Manufacturing Liang Zhang, University of Connecticut, 371 Fairfield Way, Unit 4157, Storrs, CT, 06269, United States, liang@engr.uconn.edu, Zhiyang Jia Many manufacturing systems operate under finite production runs. In such systems, a certain number of products (e.g., a customer order) are entered into the system. The in-process buffers are typically empty when a production run starts. This leads to transient period to build up the throughput and inventory during the process. One of the key problems for such systems is to calculate the performance of the system while completing such a production run. In this abstract, we discuss analytical and approximate methods to accomplish this. 2 - Management of a Shared Spectrum Network in Wireless Communications – A Queueing Approach Shining Wu, Assistant Professor, The Hong Kong Polytechnic University, LMS.Dept, Hong Kong, 999077, Hong Kong, sn.wu@polyu.edu.hk, Jiheng Zhang, Rachel Q. Zhang We consider a band of the electromagnetic spectrum with a finite number of identical channels shared by both licensed and unlicensed users. Such a network differs from most many-server, two-class queues in service systems including call centers due to the restrictions imposed on the unlicensed users in order to limit interference to the licensed users. We first approximate the key performance indicators under the asymptotic regime and then study the optimal sharing decisions of the system to maximize the system throughput rate while maintaining the delay probability of the licensed users below a certain level. Our analysis reveals a number of distinctive properties of the system. 3 - Online Priority Service Pricing with Heterogeneous Customers Ping Cao, University of Science and Technology of China, University of Science and Technology of China, Hefei, 230026, China, pcao@ustc.edu.cn, Jingui Xie, Yaolei Wang Consider a service system with a single server and two queues (i.e., the VIP queue and the ordinary queue), in which the server gives priority to the VIP queue. Customers are heterogeneous in their waiting cost rate, the distribution of which is unknown. If the arriving customer joins the VIP queue, he will be charged a fee K. The service system’s objective is to maximize his long-run average profit by setting the additional fee. We find that there might be multiple equilibria of customers’ chosen behavior under a fixed fee and the system can always reach the largest equilibrium by setting a proper initial delay announcement. Later, we use an online optimization method to find the optimal charged fee. 4 - New Perspective on Performance Measurement in Multi-server Queuing Systems Hung Tuan Do, University of Vermont, 55 Colchester Avenue, 207 Kalkin Hall, Burlington, VT, 05405, United States, hdo@uvm.edu, Masha Shunko, Alan Scheller-Wolf In some service systems, performance in the right tail of the queue size and/or waiting time distribution is the main source of managerial concern. We propose a new perspective on analyzing performance of such systems by defining new stochastic orders to compare the tail of distributions and performance measures based on these orders. We also propose a class of control policies that guarantee Pareto improvement based on the newly dened measures. WA62

3 - Optimal Pairs Trading with Dynamic Mean-variance Dongmei Zhu, Southeast University, Nanjing, China, dongmeizhu2013@126.com

Pairs trading strategy has been popularly adopted by hedge funds.In this paper, we investigate optimal pairs trading strategies for dynamic trading problems.The price difference of a pair of securities is modeled by an Ornstein-Uhlenbeck (OU) process.We study and formulate the mean-variance (MV) portfolio problem and the optimization problem is solved dynamically by utilizing the purely-buy-and-

464