Ananth V. Iyer

Improved Fashion Buying with Bayesian Updates

We focus on the problem of buying fashion goods for the “big book” of a catalogue merchandiser. This company also owns outlet stores and thus has the opportunity, as the season evolves, to divert inventory originally purchased for the big book to the outlet store. The obvious questions are: (1) how much to order originally, and (2) how much to divert to the outlet store as actual demand is observed. We develop a model of demand for an individual item. The model is motivated by data from the womens designer fashion department and uses both historical data and buyer judgement. We build a stochastic dynamic programming (DP) model of the fashion buying problem that incorporates the model of demand. The DP model is used to derive the structure of the optimal inventory control policy. We then develop an updated Newsboy heuristic that is intuitively appealing and easily implemented. When this heuristic is compared to the optimal solution for a wide variety of scenarios, we observe that it performs very well. Si...

On an edge ranking problem of trees and graphs

Abstract A k-edge ranking of an undirected graph is a labeling of the edges of the graph with integers 1, 2, …, k, with the property that all paths between two edges with the same label i contain an edge with label j[rang]i. The edge ranking problem is that of finding the smallest k for which a graph has a k-edge ranking. This problem is useful in the optimization of the number of parallel stages required to assemble a product from its components. The problem is also related to that of finding minimum height edge partition trees of graphs. The main result in the paper is an O(n log n) time approximation algorithm for edge ranking of trees, which has a worst case performance ratio of 2.

Assessing the Value of Information Sharing in a Promotional Retail Environment

We focus on a logistics system where inventory is held at three levels: the customers, the retail store, and the warehouse. Retail customer segments are heterogeneous and differ in their reservation prices for product as well as their holding costs. They purchase product from a retail store managed by a retailer. The retailer chooses a retail pricing scheme to maximize his expected profit given a model of customer temporal response to retail pricing. This retailer is supplied product from a warehouse managed by a manufacturer.The manufacturer is responsible for maintaining inventory level at the warehouse and providing 100% service level for retailer orders. The manufacturer uses all available information to generate an inventory policy that maximizes expected profit subject to the service-level requirement. We evaluate the manufacturers optimal expected profit under two possible schemes: (1) no information regarding the timing of retail promotion plans, and (2) full information regarding the timing of retail promotion plans.We show: (1) as the predictability of the sales impact of a promotion decreases, it may be optimal for the retailer to eliminate retail promotions; (2) increased stockpiling tendency of customers increases retailer profits and decreases manufacturer profits; and (3) retail-promotion information sharing can make retail promotions change from being less profitable than no promotions to being more profitable than no promotions for the manufacturer. We show the impact of fitting the model to a grocery store data set that provided data regarding retail sales (and associated prices) of canned tomato soup over two years. We also explore managerial insights suggested by the model.

Information networks and market behavior

This paper proposes a mathematical model of financial markets as networks. The model examines the effect of network structure on market behavior (price volatility and trading volume). In the model, investors are arrayed in various network configurations through which they gather information to make trading decisions. The basic network considered is a chain graph with two parameters, number of investors (n) and the length of time in which information is transmitted (k). Closed‐form expressions for price volatility and expected trading volume are provided. The model is generalized to more complex networks, focusing on the hub‐and‐spoke network. The network configurations analyzed do not represent the real (and unknown) communication network among investors, but predictions from the model are consistent with price and volume patterns observed in sociological and economic research on financial markets. The main result is that network structure alone influences price volatility and expected trading volume even...

Indian Auto-Component Supply Chain at the Crossroads

We trace the evolution of the auto-component supply chain in India beginning with the opening of the economy in 1990 by using a combination of data on firm and sector performance, customer-satisfaction surveys, and interviews with experts. During the past decade, the industry has made remarkable progress on multiple fronts. This is particularly true with regard to quality---10 firms in this industry have won the coveted Deming prize during the past six years. Surprisingly, we first observe that the financial performance of the firms that won the Deming prize (i.e., Deming firms) shows no definitive differences from the performance in the rest of the industry. We then analyze the productivity growth at the firm level across two five-year intervals using a total-factor productivity model. Our results suggest that productivity improved much more during the second period, which is the interval in which most of the firms won the Deming prize. We also analyze the impact of winning the award on profitability and suggest that new firms were able to grow faster in the improving business environment. To “externally” validate our findings, we compare the auto sector in India with that in China. Despite a 10-year disadvantage because of costs that are beyond the control of the firm, the auto sector in India seems to be competitive with that sector in China on all firm-specific factors. In summary, we suggest that firms in this sector have taken the first step by becoming competitive in the areas of cost and quality. We suggest that they are now at a crossroads and must make several choices to leverage these quality gains into a profitable, global supply chain strategy.

Efficient Supply Chain Management at the U.S. Coast Guard Using Part-Age Dependent Supply Replenishment Policies

The United States Coast Guard (USCG), now part of the Department of Homeland Security, has the mission to secure the U.S. coastline using a combination of air and sea capabilities. This paper focuses on an application of operations research techniques at the USCG to improve the performance of its aircraft service parts supply chain. We focused on evaluating the supply chain benefits from linking the aircraft maintenance database with the aircraft parts inventory database. This required us to (a) develop an approach to link the databases and (b) use aircraft maintenance information to improve the inventory management of service parts at the USCG. We first used mathematical programming tools to merge the maintenance database with the demand database. We then developed state-dependent supply replenishment policies that use part-age information to manage the service parts supply chain. We show that one of the proposed policies permits analytic estimation of the benefits of linking the data sets. The impact of these inventory policies was evaluated using empirical demand data for 41 critical parts over a five-year period. Computational results suggest that our proposed policies can lead to significant reductions in inventory cost over the current system, as high as 70% for some parts. Based on the insights from this study, the USCG is currently contracting with commercial vendors to develop an operational database and decision-support implementation across all parts.

Contingency management under asymmetric information

We model a monopolist supplier whose supply to multiple buyers is disrupted. The supplier can take costly, speed-dependent actions, to restore supply. Buyers experience private backorder costs that are unknown to the supplier. We analyze the suppliers optimal contract structure and explore the impact of an alternate supplier.

An Approach to Securely Identifying Beneficial Collaboration in Decentralized Logistics Systems

The problem of sharing manufacturing, inventory, or capacity to improve performance is applicable in many decentralized operational contexts. However, the solution of such problems commonly requires an intermediary or a broker to manage information security concerns of individual participants. Our goal is to examine use of cryptographic techniques to attain the same result without the use of a broker. To illustrate this approach, we focus on a problem faced by independent trucking companies that have separate pick-up and delivery tasks and wish to identify potential efficiency-enhancing task swaps while limiting the information they must reveal to identify these swaps. We present an algorithm that finds opportunities to swap loads without revealing any information except the loads swapped, along with proofs of the security of the protocol. We also show that it is incentive compatible for each company to correctly follow the protocol as well as provide their true data. We apply this algorithm to an empirical data set from a large transportation company and present results that suggest significant opportunities to improve efficiency through Pareto improving swaps. This paper thus uses cryptographic arguments in an operations management problem context to show how an algorithm can be proven incentive compatible as well as demonstrate the potential value of its use on an empirical data set.

The Supply Chain Impact of Smart Customers in a Promotional Environment

Increasing product variety through the use of alternate package sizes is a commonly observed mechanism in the grocery industry. Under such a scheme, however, the response to pricing decisions for each of the different package sizes is affected by how customers make demand choices. We build a demand model in which customers reactsmart to retail promotions through stockpiling and package size switching. The demand model combines a customer choice model with a model in which customers differ in their stockpiling and reservation price levels. We utilize data from the German grocery industry for an empirical fitting of the model. We then develop a store-level inventory model for each SKU and optimize price promotions to maximize expected profit. We show the benefit of capturing thesmart customer response to price promotions by demonstrating its impact on the reduced inventory costs. We use the model to generate a number of managerial implications of the model for the German grocery environment.

The Logistics Impact of a Mixture of Order-Streams in a Manufacturer-Retailer System

We model a supply chain with two retail warehouses that place replenishment orders with a common manufacturing capacity. The two retailers differ in the variability of their order-streams. The order-stream from one retail warehouse is modeled as a Poisson process and from the other as a hyperexponential renewal process. Each retail warehouse uses a base-stock policy to place replenishment orders with the manufacturer. The manufacturer is modeled as a first-come-first-serve, single exponential server queue. We analyze the supply-side impact of this mixture of order-streams received by the manufacturer on both retailers. An exact analysis of this base-model generates closed-form expressions for distributions of the lead-time, outstanding orders, and expected inventory costs for each retailer, and leads to comparative results about the two retailers- performance measures. The base-model is extended to accommodate finished goods at the manufacturer, more than two retailers, and bulk-arrivals. We use the model to suggest managerial insights about the impact of the presence of a high-variability retailer on other retailers who share capacity, the distorting impact of manufacturer finished goods inventory on retailer incentives, and the incentives for retailers to participate in variability-reduction programs in the grocery industry.