Informs Annual Meeting 2017

TE31

INFORMS Houston – 2017

TE29

6 - Controlling Endogeneity in Event Studies: Valuing Movie Product Placements using Prediction Market Data David Richardson, Assistant Professor of Marketing, Illinois Institute of Technology, Chicago, IL, 60605, United States, david.richardson@stuart.iit.edu Abstract: Event studies are widely used in marketing to value intangible investments using abnormal returns as measured by equity markets. Two objections; omitted variables which are endogenously related to event timing may bias estimates, and researcher flexibility in choosing the event window allows for cherry picking results. We use data from the Hollywood Stock exchange (HSX) to control for these effects in a study of product placement of automobiles in films. We evaluate individual placements using regression on panel data and compare the results with the conclusions from a standard event study approach to shed light on potential endogeneity bias in such studies. 351A Finance Contributed Session Chair: Nadja Younes, University of Konstanz, Konstanz, Germany, nadja.younes@uni-konstanz.de 1 - Analyst Coverage, Board Monitoring, and Firm Value. Evidence from a Natural Experiment Torsten Twardawski, University of Konstanz, Universitaetsstr. 10, Konstanz, 78467, Germany, Torsten.twardawski@uni-konstanz.de, Iana Zborshchyk In this paper, we estimate a causal impact of analyst coverage on the long-term value of a company. Using two established quasi-natural experiments, we find that exogeneous coverage reductions decrease long-term firm value. We show that the reduction in firm value is linked to a decline in fundamental value and stronger if a board has weak monitoring power. Conditional on their initial internal monitoring quality, firms try to offset the negative effect by improving their internal governance mainly through getting more independent boards. Overall, our results confirm that analysts create value through their governance role, which interacts with preexisting internal monitoring. 2 - The Role of Sentiment-driven Institutional Investors in IPO Pricing Bo Liu, University of Electronic Science and Technology of China, No. 2006, Xiyuan Ave, West Hi-Tech Zone, Chengdu, 611731, China, b.liu07@fulbrightmail.org, Jingbin He In this paper, we take advantage of the unique setting and big data of Chinese auction-based IPO process to examine the role that sentiment institutions play in the pricing of newly listed stocks. We find that underwriters may “strategically” use positive/negative language to manipulate the writing style in the premarket Valuation Report to influence bidder’s pricing and hence the final offer prices of IPO. This is one type of “agency problem”, which leads to substantial losses of both retail and institutional investors and more volatile post-IPO stock prices. The results suggest that sentiment institutional investors deteriorate the pricing of newly listed stocks. 3 - Financial Assets and Demand Estimation Murat Tiniç, PhD Candidate, Bilkent University, Ankara, Turkey, tinic@bilkent.edu.tr In this study, I propose to model the investment decisions of an investor as a discrete choice action, and estimate the demand for a financial asset through the underlying structure of the investor’s utility. 4 - The Impact of a CEO’s Status as High Quality Board Director , on Personal and Organizational Outcomes: Evidence fro CEO Awards Nadja Younes, University of Konstanz, Universitaetsstr. 10, Konstanz, 78467, Germany, nadja.younes@uni-konstanz.de, Torsten Twardawski In this paper we examine the extent to which a CEO’s status as high quality board director impacts economic outcomes. We use a unique data set of CEO awards to measure positive second-order status shocks to CEOs, who serve on an award- winning CEO’s board. Around the awarding, we find negative and significant stock-price reactions for the CEOs’ focal firms. We show that affected CEOs accept more board positions and receive higher compensation. Corporate governance has a moderating effect on the investigated outcomes. Our findings suggest that CEOs with reputable board positions become busy with firm-unrelated activities, dedicate less time and effort to their firms, and destroy shareholder wealth. TE31

350E Manufacturing Contributed Session Chair: David Richardson, Illinois Institute of Technology, Chicago, IL, United States, david.richardson@stuart.iit.edu 1 - Measuring Large-scale Market Responses from Aggregated Sales: Regression for High-dimensional Sparse Data Nobuhiko Terui, Professor, Tohoku University, Kawauchi Aoba-Ku, Sendai, 980-8576, Japan, terui@econ.tohoku.ac.jp We propose a regression model for high-dimensional sparse data from store-level aggregated sales. The modeling comprises two sub-models—topic model and hierarchical factor regression model—that are applied sequentially not only for accommodating high dimensionality and sparseness but also for managerial interpretation. First, the topic model is applied unusually to aggregated data to decompose the aggregated sales volume of a product into sub-sales for several topics by allocating each unit sale in a day into one of topics based on joint purchase information. Then, we construct a topic-wise market response function by hierarchical factor regression based on canonical correlation analysis. 2 - In Store Customer Experience Creation in the Retail Industry Trisha D. Anderson, Texas Wesleyan University, 1201 Wesleyan Street, Fort Worth, TX, 76105, United States, trdanderson@txwes.edu This paper examines the impact of the right mix of atmospheric factors used to create the in-store customer experience in a retail market that includes ambience factors (volume of music, type and strength of aroma, level of lighting), interaction with sales people, and store display, on the in-store customer experience, the customer’s emotional state, and customer shopping behavior. 3 - Do Marketing Talents Matters its Relationship to Marketing Capability and Performance YiChun Liu, National Taiwan University, Room 1116, No 30, Sec 3, Taipei, 10668, Taiwan, d02724003@ntu.edu.tw The discussion of marketing capability on firms’ performance has been well documented for the last two decades. Most of the previous studies regarding the measurement of marketing capability are based solitary on the attitude of top managers. However, employees’ marketing talents are rarely discussed. This study investigates a new construct of marketing capability and the measurement of individual employee’s marketing talents. A preliminary questionnaire was developed to conduct an empirical study, of which over 2,000 employees were surveyed. This study identified the positive impact of employee marketing talent Launching a new program is challenging for organizations because their potential clients have to learn its unknown quality. This is the first paper to empirically study how adoption by high school classmates and coworkers affects each peer’s adoption of a new employment subsidy during implementation stages. Identification comes from a discontinuity in the subsidy assignment rule inducing exogenous variation in a neighborhood around the worker’s wage eligibility cutoff. Using a unique set of administrative records, we find that: coworkers strongly influence one’s adoption, classmates do not; peer effects decrease over time. Hence, information diffusion can explain adoption in the short run. 5 - The Power of Social Networks in Predicting Noncontractual Customer Activity Evidence from Venmo Pantelis Loupos, PhD Candidate, Kellogg, Northwestern University, 2800 N Lakeshore Dr, Apt 2913, Chicago, IL, 60657, United States, louposp@gmail.com There is a shift away from the subscription-based business model to on-demand (OD) services. We focus on predicting customer activity (CA) in OD settings in the presence of social network (SN) data. Data comes from Venmo, a P2P payment application. We: 1) Conduct a dynamic analysis to quantify the effect of user and SN metrics on CA and find that SN metrics help achieve a significant increase in predictive accuracy during the beginning of a customer’s lifetime. 2) Evaluate our models in the task of CA-based customer segmentation and find that we can identify customers segments with high accuracy. 3) Identify structural differences among the different customer segments. and managerial marketing capability to organizational performance. 4 - Peer Effects in the Adoption of a New Social Program Claudio Mora, Pontificia Universidad Javeriana, Carrera 7 #40b-36, Piso 4, Bogota, Colombia, claudio_mora@javeriana.edu.co

421