Research Seminars 2014/2015

Aurelie THIELE - Lehigh University, USA

May 18th 2015

Robust Design of (American) Health Insurance Plans

Abstract: We investigate optimization models to design a menu of health insurance plans offered by a large (American) employer self-insuring through a private health exchange, in the presence of uncertainty in employees’ choice of plans, health status and health resource utilization. While the application of this talk is grounded in the American healthcare landscape, it incorporates risk management, choice models, robust optimization, insurance concepts and, more broadly, decision-making under high uncertainty and customer choice. Our goal is to optimize, subject to a budget constraint on the employer’s part, an employee-driven criterion such as the minimization of employees’ out-of-pocket costs or the maximization of the fairness of the design, with a focus on healthcare cost as a fraction of employee’s income. We analyze the employer’s cost/long-term risk trade-off and analyze policy choices to promote employees’ health while maintaining premiums at a sustainable level. Of key interest is the allocation of the employer’s budget to various health care costs, such as prescription drugs or hospital stays, to achieve health objectives for the employee population without waste. We consider both traditional health plans and High-Deductible Health Plans (HDHPs), which give employees an incentive to be cost-conscious in occurring health expenses below the deductible – with the exception of preventive services required to be offered for free by the plan – but may also delay care. We provide an heterogeneous view of the employee population based on the type of plan employees select (from Health Maintenance Organizations with a local provider network and a requirement to have a referral to access the specialty-care network, to more expensive Preferred Provider Organizations that do not require referrals to make appointments with specialists) and plans’ actuarial value or metal level (bronze, silver, gold and platinum having actuarial values of 60%, 70%, 80% and 90%, respectively, which represents the part of healthcare costs the payer is expected to shoulder for a benchmark enrollee population). In addition, we discuss how to select the optimal number of plans on offer in the private health exchange. Joint work with Dimitris Bertsimas and Jerry Chen of MIT.

Felix PAPIER - ESSEC Business School, France

April 13th 2015

How to Split the Pie ? Sequential Supply Allocation with Forecast Updates

Abstract: We study the problem of allocating supply under advance demand information (ADI). We consider a company that must allocate limited inventory to different markets that open sequentially. To reduce uncertainty, the company receives advance demand information and updates forecasts about its markets each time it makes an allocation decision. We study the value and optimal use of this information. Our research is motivated by an agri-food manufacturer that operates in several European countries. We develop the optimal policy under relaxed conditions and an efficient heuristic policy that performs close to optimal under general conditions. We derive structural properties of the model to gain managerial insights, and we derive the optimal policy in closed-form for the case of markets with identical prices. We use numerical experiments to demonstrate that the value of ADI can be significant. The managerial insights of this study include the observations that in environments as the one that motivated our research, early markets receive systematically less supply than late markets, and that the value of ADI is greatest if the initial supply is close to the initial forecasts.

Francis DE VERICOURT - EMTS Berlin, Germany

March 2nd 2015

Financing Capacity Investment under Demand Uncertainty

Abstract: This paper studies the interplay between the operational and financial facets of capacity investment. We consider the capacity choice problem of a firm with limited liquidity and whose access to external capital markets is hampered by moral hazard. The firm must therefore not only calibrate its capacity investment and the corresponding funding needs, but also optimize its sourcing of funds. Importantly, the set of available sources of funds is derived endogenously and includes standard financial claims (debt, equity, etc.). We find that when higher demand realizations are more indicative of high effort, debt financing is optimal for any given capacity level. In this case, the optimal capacity is never below the efficient capacity level but sometimes strictly above that level. Further, the optimal capacity level increases with the moral hazard problem's severity and decreases with the firm's internal funds. This runs counter to the newsvendor logic and to the common intuition that by raising the cost of external capital and hence the unit capacity cost, financial market frictions should lower the optimal capacity level. We trace the value of increasing capacity beyond the efficient level to a bonus effect and a demand elicitation effect. Both stem from the risk of unmet demand, which is characteristic of capacity decisions under uncertainty. Joint work with Denis Gromb.

Oualid JOUINI - Ecole Centrale Paris, France

February 16th 2015

Call Centers with Delay Information: Models and Insights

Abstract: We analyze a call center with impatient customers. We study how informing customers about their anticipated delays affects performance. Customers react by balking upon hearing the delay announcement, and may subsequently renege, particularly 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. Modeling the call center as an M/M/s+M queue with endogenized customer reactions to announcements, we analytically characterize performance measures for this model. The analysis allows us to explore the role announcing different percentiles of the waiting time distribution, i.e., announcement coverage, plays on subsequent performance in terms of balking and reneging. Through a numerical study we explore when informing customers about delays is beneficial, and what the optimal coverage should be in these announcements. It is shown how managers of a call center with delay announcements can control the tradeoff between balking and reneging, through their choice of announcements to be made.

Mohammed ABDELLAOUI- HEC-Paris/CNRS, France

December 15th 2014

Recursive Rank-dependent Utility for Ambiguity

Abstract: This paper proposes an ambiguity model that accounts for Ellsberg-and-Allais type behavior in the famous Anscombe and Aumann framework. Ambiguity attitudes are captured through both utility (as in recursive expected utility) and non-additive probabilities (as in Choquet expected utility), hence, combining `smooth' and `kinked' approaches to ambiguity. The model is based on a single preference principle called substitution consistency. Due to a natural embedding of backward induction, substitution consistency allows for one-stroke representations of preferences including Schmeidler's Choquet expected utility, recursive expected utility and subjective expected utility as particular cases, without referring to mixtures of lotteries. In addition to the provision of a unified setup in which different ambiguity models can jointly be analyzed and compared, substitution consistency simplifies the formal study of relative concavity of utility for risk and ambiguity without committing to recursive expected utility. We also show how our general recursive model can facilitate the descriptive study of ambiguity attitudes. Joint work with Horst Zank (University of Manchester).

Gila E. FRUCHTER - Bar-Ilan University, Israel

November 24th 2014

Problem-Driven Theory with Theory-Driven Solutions

Abstract: In my research in marketing I seek to develop a problem-driven theory and find theory-driven solutions. In finding such solutions I draw on my long-standing experience of optimal control. I chose a closer look into two of my recent publications to have a taste of my work. We start with dynamic pricing for subscription services, continue with production location decisions of brands.

Phd Thesis Defense of affiliated cluster member Cerasela TANASESCU - ESSEC Business School, France

November 5th 2014, 10h, room N517, presentation in French

Georges ZACCOUR - HEC Montréal, Canada

November 3rd 2014

Optional-Contingent-Products Pricing in Supply Chains

Abstract: This paper studies the pricing strategies of firms belonging to a vertical channel structure where optional contingent products are sold. Optional contingent products are characterized by unilateral demand interdependencies. That is, the base product can be used independently of a contingent product. On the other hand, the contingent product's purchase is conditional on the possession of the base product. We find that the retailer decreases the price of the base product to stimulate demand on the contingent-product market. Even a loss-leader strategy could be optimal, which happens when reducing the base product's price has a large positive effect on demand of this base product, and thus on the number of potential consumers of the contingent product. The price reduction of the base product also mitigates the double-marginalization problem, which is well known in a supply-chain setting with one manufacturer and one retailer, in a large part of the parameter space. Joint work with P.M. Kort (Tilburg University), and S. Taboubi (GERAD, HEC Montréal).

Steffen JORGENSEN - University of Southern Denmark, Denmark

October 6th 2014

Recent Development in Lanchester Differential Games.

Abstract: Lanchester games have their origin in the work of F.W. Lanchester (F.W. Lanchester: Aircraft in Warfare: The Dawn of the Fourth Arm. Constable, London, UK, 1916) who used ordinary differential equations to study stylized problems of military combat. Later on, one of Lanchester’s models has been used - quite successfully - to describe the effects of advertising competition in oligopolistic markets. A stream of literature in the field of advertising has used this model as a component of a dynamic game. I introduce a couple of Lanchester’s models and a simple dynamic advertising game model. The advertising game has been extended in various directions, for example, to take into account market growth/contraction and multiple types of advertising efforts. Such extensions are discussed. I present two examples of my own current research in the area and conclude by suggesting some avenues for future research.