School of Business
rss twitter LinkedIn Facebook YouTube Blogs

Past Seminars in Spring 2009:

Title: Examining Product Modularity from a Behavioral Perspective: Do Consumers Properly Value Modularity?

Speaker: Sezer Ulku
McDonough School of Business
Georgetown University


In a variety of fields, from strategy to operations management and engineering design, product modularity has been advocated as a design strategy, providing many benefits for customers, firms and the environment. From a customer perspective, modularity would allow adaptation of products in line with technological improvements and changing needs. From the firm's perspective, modularity can translate into lower design and production costs. From an environmental perspective, modularity has the potential to reduce the consumption of natural resources and pollution by extending the useful life of product components. Despite these benefits, modular products, as experienced by customers, are relatively uncommon Motivated by this observation, we examine the customer perspective: Through behavioral experiments, we explore the biases and heuristics that shape customer's valuation of modular products, and their purchase and upgrade decisions. Among the many facets of modularity, we focus on modular upgradeability, which allows the replacement of components, instead of whole product to improve product performance. First, we examine the choice between modular and integrated products: When is a modularly upgradeable product viewed favorably by consumers? How do time between generations, and uncertainty in pricing and technology impact the willingness to pay for a modular platform? Next, we consider products where the customer already owns a modular product (and therefore has the option to upgrade), and examine under which circumstances the upgrade option is exercised. We find that consumers may undervalue (overvalue) modular upgradeability for rapidly (slowly) improving products relative to rational economic value due in line with hyperbolic discounting. When compared to decisions involving cash, hyperbolic discounting is more pronounced in making product choices. Furthermore, when there is uncertainty or replacement effort, the value of modularity is further reduced. These findings provide partial demand-side explanation to why there are few products that are modularly upgraded. Finally, we observe that by providing support or a donation opportunity, customers can be encouraged to choose modular upgrades.

Time: Friday, May 8th 3:30 - 4:30 pm

Location: Duques Hall 552

Title: Introduction to Risk-Averse Optimization

Speaker: Andrzej Ruszczynski


We shall discuss several issues associated with modeling risk aversion in stochastic optimization. The first part of the talk will be devoted to mean-risk models and general risk functionals. We present duality theory for such functionals and investigate their consistency with stochastic orders. Next, we develop optimality conditions and duality theory for optimization problems involving risk functionals. The concept of risk value of perfect information will be introduced as well. Then we pass to constructing risk functionals for random reward sequences. For this purpose we introduce the concept of a conditional risk mapping, analyze its properties, and develop duality relations. Finally we mention dynamic optimization problems involving such functionals and we present optimality conditions of dynamic programming type. The second part of the talk will be devoted to modeling risk aversion by stochastic dominance constraints. We consider optimization problems with such constraints and present optimality conditions and duality theory for these problems. We show that utility functions play the role of Lagrange multipliers associated with stochastic dominance constraints. We also mention combinatorial questions associated with first order dominance constraints and extensions to dynamic problems. Finally, we have a financial planning example.

Time: Friday, May 1st 3:30 - 4:30 pm

Location: Duques Hall 553

Title: Bootstrap Tests of Stationarity

Speaker: Tara M. Sinclair
Department of Economics and the Elliott School of International Affairs
The George Washington University


We compare the finite-sample performance of different stationarity tests. Monte Carlo analysis reveals that tests based on Lagrange multiplier (LM) statistics with nonstandard asymptotic distributions reject far more often than their nominal size for trend-stationary processes of the kind estimated for macroeconomic data. Bootstrap versions of these LM tests have empirical rejection probabilities that are closer to nominal size, but they still tend to over-reject. Meanwhile, we find that a bootstrap likelihood ratio (LR) test has very accurate finite-sample size, while at the same time having higher power than the bootstrap LM tests against empirically relevant nonstationary alternatives. Based on the bootstrap LR test, and in some cases contrary to the bootstrap LM tests, we can reject trend stationarity for US real GDP, the unemployment rate,consumer prices, and payroll employment in favour of unit root processes with large permanent movements.

Time: Friday, April 24th 11:00 am - 12:00 noon

Location: Duques Hall 453

Title: The Impact of Supply Quality and Supplier Development on Contract Design

Speaker: Sila Cetinkaya
Department of Industrial and Systems Engineering
Texas A&M University


In this talk, we examine two key issues in supply management: supply quality and supplier development. To this end, we consider a supplier-buyer pair and develop analytical models for designing optimal buyer-initiated supply contracts with lot sizing, supply quality, and supplier development considerations while modeling private information and individual incentives explicitly. We study two distinct contractual settings. First, we concentrate on the case where there is no supplier development, and, hence, no supply quality improvement effort. In this case, we show that the contractual lot size is larger than the channel optimum when the buyer has incomplete information about the supplier's quality level. For the special case where the buyer's prior distribution of supplier's quality level is uniform, we prove that immediate contracting is in fact more efficient for the channel than not having a contract in the long run as long as the buyer's prior expected quality level is sufficiently high, i.e., more than 75%, or the buyer's estimation of the quality level is unbiased. However, for general prior distributions of the supplier's quality level, immediate contracting may not be effective for the channel depending on the characteristics of the hazard rate function of the prior distribution. Since the efficiency of the contract depends on the buyer's prior distribution of the supplier's quality level, we also present a dynamic programming model for the buyer to determine when to offer a contract to the supplier under information updating. Next, we concentrate on the case where the buyer seeks a quality improvement initiative under a supplier development program but she has incomplete information about the supplier's quality investment sensitivity. We show that the buyer will request a lower level of quality improvement than in the full information case. Also, in this case, we demonstrate that buyer-initiated contracting under asymmetric information is always worthwhile; however, contracting may not lead to quality improvement. In particular, depending on the characteristics of the reverse hazard rate function of the buyer's estimation of the supplier's investment sensitivity, investment decision may not be made. As a result, information asymmetry may ruin the buyer's interest in initiating a supplier development program in practice.

Time: Friday, April 17th 11:30 am - 12:30 pm

Location: Duques Hall 453

Title: Delay Announcements in Call Centers

Speaker: Zeynep Aksin-Karaesmen
Koc University and Northwestern University


Announcing delays to customers who are put on hold in a call center is a common practice. Managers can have different objectives in providing such information: modulating demand by signaling times of high congestion, enhancing satisfaction with inevitable waiting, or both. These objectives bring with them several challenges: estimating real-time delays for each customer in a stochastic environment and deciding on what to announce to these customers. In this paper we model customer reactions to delay announcements. We study how informing customers about their anticipated delays affects performance. Customers react by balking upon hearing the delay announcement, and may subsequently renege if the realized waiting time exceeds the delay that has originally been announced to them. The balking and reneging from such a system are a function of the delay announcement coverage. Modeling the call center as an M/M/s+M queue, we analytically characterize performance measures for this model, and using these within a numerical study we explore when informing customers about delays is beneficial, and what the optimal coverage should be in these announcements.

Time: Friday, March 27th 11:30 am - 12:30 pm

Location: Duques Hall 453

Title: Sequential Predictive Regressions and Optimal Portfolio Returns

Speaker: Nicholas Polson
Professor of Econometrics and Statistics
The University of Chicago Graduate School of Business


This paper analyzes sequential learning in the context of predictive regression models. To do this, we develop new particle based methods for sequential learning about parameters, state variables, hypotheses, and models. This sequential perspective allows us to quantify how investor's views about predictability and models varies over time, and naturally mimics the learning problem encountered in practice. We consider learning about predictability using dividend/payout data and models that incorporate drifting coefficients and stochastic volatility. We analyze the time-variation of parameter estimates and model probabilities, using both the traditional cash dividends measure and a measure taking into account share repurchases and issuances. We also analyze the economic benefits of using these models by considering optimal portfolio allocation problems.

Time: Friday, March 13th 11:00 am - 12:00 noon

Location: Duques Hall 552

Title: Combinatorial Patterns for Probabilistically Constrained Optimization Problems

Speaker: Miguel Lejeune
Department of Decision Sciences
The George Washington University


We propose a new framework for the solution of probabilistically constrained optimization problems by extending some recent developments in combinatorial pattern theory. The method involves the binarization of the probability distribution and the generation of a consistent partially defined Boolean function (pdBf) representing the combination (F,p) of the binarized probability distribution F and the enforced probability level p. We represent the pdBf representing (F,p) as a disjunctive normal form taking the form of a collection of combinatorial patterns. We propose a new integer programming-based method for the derivation of combinatorial patterns and present several methods allowing for the construction of a disjunctive normal form that defines necessary and sufficient conditions for the probabilistic constraint to hold. The obtained disjunctive forms are then used to generate deterministic reformulations of the original stochastic problem. The method is implemented for the solution of a numerical problem. Extensions to the present study are discussed.

Time: Friday, March 6th 3:30 - 4:30 pm

Location: Duques Hall 552

Title: Supply Chain Efficiency and Contracting in the Presence of Gray Market

Speaker: Mehmet Sekip Altug
Graduate School of Business
Columbia University


In recent years, a growing number of companies have witnessed a new form of supply chain developing for their products outside their own regular channel. Known as gray market or "parallel" channels, the products sold through these supply chains are genuine brand owner's products, yet their sales are completely unauthorized by the brand manufacturer. We consider such a supply chain with one manufacturer and several authorized retailers that face uncertain demand and a potential gray market. While the gray market can be seen as an opportunity to sell any excess inventory, it also is a threat for authorized retailer's demand. We characterize the equilibrium market-clearing gray market price and derive a closed-form result for a special case which approximates the equilibrium gray market price very accurately. We determine how the equilibrium changes with respect to model parameters such as the quality of the gray market product, wholesale price and heterogeneity of the consumer valuation. Assuming that the manufacturer takes a naive approach and charges the same wholesale price as if no gray market exists, we derive the conditions under which gray market could be a threat or an opportunity. We then characterize manufacturer's wholesale pricing problem when it is strategic and anticipates the gray market emergence. Comparing decentralized and centralized system in a gray market environment, we show that wholesale pricing contract itself is "almost" coordinating. Finally, we analyze several other contracts that are designed not taking into account the potential for gray market and answer two main questions when these contracts are in use and a gray market emerges in the end: i) How do the gray market dynamics change under these contracts? ii) From manufacturer's standpoint, how does the performance of these contracts compare to one another? Based on our results, we derive several managerial insights and position gray market as an external business factor that could provide one explanation for the commonality of certain contracts in industry.

Time: Friday, February 27th 11:30 am - 12:30 pm

Location: Funger Hall 453

Title: Bayesian Spatial Prediction

Speaker: Benjamin Kedem
Department of Mathematics
University of Maryland


We discuss Bayesian spatial/temporal prediction in transformed Gaussian random fields where the transformation belongs to a parametric family. Monte Carlo integration is used in the approximation of the predictive density function, which is easy to implement in this framework. The BTG software for the implementation of the method will be discussed by means of spatial and temporal examples. As a byproduct, we provide a Bayesian way to tackle the distribution problem of average rainfall rate.

Time: Friday, February 20th 3:30 - 4:30 pm

Location: Duques Hall 552

Title: Random Partition Models Indexed with Covariates

Speaker: Peter Mueller
M.D. Anderson Cancer Center


We propose a model for covariate-dependent clustering, i.e., we develop a probability model for random partitions that is indexed by covariates. The motivating application is inference for a clinical trial. As part of the desired inference we wish to define clusters of patients. Defining a prior probability model for cluster memberships should include a regression on patient baseline covariates. We build on product partition models (PPM). We define an extension of the PPM to include the desired regression. This is achieved by including in the cohesion function a new factor that increases the probability of experimental units with similar covariates to be included in the same cluster. We discuss an application to clinical trial design. The proposed model is used to implement borrowing of strength across nonexchangeable sub-populations.

Time: Friday, February 13th 3:30 - 4:30 pm

Location: Duques Hall 550

Title: Expectations, Disappointment, and a Behavioral Risk-Value Model

Speaker: Philippe Delquié
Fontainebleau, France


I will present a new model of decision under risk based on a simple behavioral hypothesis: in facing a risky prospect, individuals care about how they will come out relative to all prospective outcomes of the situation, not a single benchmark as is commonly assumed in defining risk measures. The resulting model yields an intuitively appealing Risk-Value representation, which includes some classic measures of risk as special cases, such as the variance or Gini mean difference, although it is distinct from the traditional families of risk measures. Necessary and sufficient conditions will be shown for the model to satisfy first and second order stochastic dominance, two essential criteria for ordering prospects. Our risk-value model explains a richer set of risk taking behaviors (compatible with empirical observations) than Rank-Dependent Utility, a most popular alternative to Expected Utility, without more degrees of freedom, and without assuming non-linear probability weighting. Results from fitting the model to experimental data will be presented. By allying behavioral realism and normative principles in a mathematically tractable fashion, the model offers a new basis for prescriptive applications in risk management, e.g., (Behavioral) Finance, Insurance, or project portfolio selection.

Time: Thursday, February 5th 11:30 am - 12:30 pm

Location: Funger Hall 520

Title: Avoiding Lawsuits with a Bayesian Approach to Product Engineering

Speaker: Robert F. Bordley
General Motors


Mathematical programming -- which involves optimizing an objective function subject to various constraints -- has long recognized that the coefficients in both the constraints and the objective function are often uncertain. Standard expected utility analysis can easily resolve problems when the uncertainties only appear in the objective function. But utility approaches have, in the past, not been considered useful when uncertainties appear in the constraints and the objective function. Instead two alternative approaches are used to these `stochastic programming problems'. The first approach treats any violation of the constraints as rectifiable in the future at some cost. Given this assumption, the stochastic optimization problem can be formulated as an unconstrained multi-stage optimization problem which can be solved with expected utility theory (even though it is not common to do so). The second approach does not assume that violation of the constraints can be rectified at some cost. This approach is widely used in, for example, reliability-based design optimization where engineers must determine design specifications for physical structures (planes, vehicles, buildings, etc.) which, if they fail to withstand certain stresses, could lead to the loss of human life. This approach maximizes an objective function subject to an upper bound (generally one in a thousand) on the probability of the constraints being violated. Unfortunately it has been shown that this approach (called chance-constrained programming), is inconsistent with utility theory and can lead to a negative value of information.

Time: Friday, January 30th 11:30 am - 12:30 pm

Location: Duques Hall 453

Title: Impacts of Carryover Parts on New Product Reliability

Speaker: Gokhan Dogan
Sloan School of Management
Massachusetts Institute of Technology


Conventional wisdom suggests that carryover parts, common parts used in successive generations of multi-generational products, have positive impacts on the reliability and durability of products because "reliability and durability of [carryover] parts have been substantially tested in the market already" (Clark and Fujimoto, 1991). However, this claim has never been tested empirically. In this paper, we make two contributions to the literature. First, by studying data from a motor vehicle manufacturer, we find that carryover parts are a major source of quality problems, contrary to conventional wisdom, as approximately half of the warranty cost is due to carryover parts. Moreover, the failure rate of carryover parts grows from one generation to the next. Second, we study ways to offset the failures of carryover parts. Using a novel simulation model, we test different policies that aim for better prioritization and analysis of carryover problems. Simulation results show that product reliability can be improved drastically using these policies. We highlight organizational and mechanical barriers to the adoption of these policies. Our results also indicate that managers should expect to witness higher warranty costs related to carryover parts on new products due to the trends in the industry.

Time: Thursday, January 29th 11:30 am - 12:30 pm

Location: Funger Hall 520

Title: An Adaptive Process Model to Support Product Development Project Management

Speaker: Tyson R. Browning
Neeley School of Business
Texas Christian University


Projects are temporary allocations of resources commissioned to achieve a desired result. Since each project is unique, the landscape between the current state (the start of the project) and the desired state (the successful end of the project) is often dynamic, uncertain, and ambiguous. Conventional project plans define a set of related activities (a work breakdown structure and activity network) with the assumptions that this set is necessary and sufficient to reach the project's desired result. Popular models for project planning (scheduling, budgeting, etc.) and control are also based on a set of project activities which are specified and scheduled a priori. However, these assumptions often do not hold, because, as an attempt to do something novel, the actual path to a project's desired result is often revealed only by the additional light provided once the work is underway. I will present a model of a product development process as a complex adaptive system. Rather than pre-specifying which activities will be done and when, we set up (1) a superset of general classes of activities, each with modes that vary in terms of inputs, duration, cost, and expected benefits and (2) simple rules for activity mode combination. Thus, instead of rigidly dictating a specific project schedule a priori, we provide a "primordial soup" of activities and simple rules through which the activities can self-organize. Instead of attempting to prescribe an optimal process, we simulate thousands of adaptive cases and let the highest-value process emerge. Analyzing these cases leads to insights regarding the most likely paths (processes) across the project landscape, the patterns of iteration along the paths, and the paths' costs, durations, risks, and values. The model also provides a decision support capability for managers. For researchers, this way of viewing projects and the modeling framework provide a new basis for future studies of agile and adaptive processes.

Time: Friday, January 23rd 11:30 am - 12:30 pm

Location: Funger Hall 320

Title: Revenue Management and Pricing: Challenges in Planning and Execution

Speaker: Itir Karaesmen-Aydin
Robert H. Smith School of Business
University of Maryland


Revenue Management (RM) has attracted significant attention from the research community in the last decade, and is cited as among the most successful use of management science and operations research tools in practice. One of the critical components in any RM application is forecasting: Initial success of RM in the travel and hospitality industries is attributed to the availability of historical information, and continuing success build on the ability to collect information. Traditional research in RM assumes availability of information about the demand, and makes assumptions about stochastic processes or probability distributions that characterize uncertainty about demand. However, models with incorrect assumptions about demand are problematic, if/when used in practice, forecasting is a persistent challenge, and, in case of new products, data is not available. Motivated by these facts, we analyze the classical multi-fare, single-leg seat inventory control problem in airline RM using limited information. Our approach employs competitive analysis of online algorithms, which guarantees a certain performance level under all possible demand scenarios. The only information required about the demand for each fare class is lower and upper bounds. We derive the best possible policies for this problem and provide extensive computational experiments and compare our methods to existing ones.

Time: Wednesday, January 21st 11:00 am- 12:00 noon

Location: Funger Hall 320