Academic Commons

Theses Doctoral

Pricing Decentralization in Customized Pricing Systems and Network Models

Simsek, Ahmet

In this thesis, we study the implications of multi-party pricing for both consumers and producers in different settings. Within most organizations, the final price of a product or service is usually the result of a chain of pricing decisions. This chain may consist of different departments of the same company as well as different companies in a specific industry. Understanding the implications of such chains on the final prices and on consumer and producer surplus is the key topic of this dissertation. In the first part of this thesis, we consider a network in which products consist of combinations of perishable resources. In this model, different revenue-maximizing "controllers" determine the resource prices and the price of the product is the sum of the prices of the constituent resources. For uncapacitated networks, we develop bounds on the "price of anarchy" -the loss from totally decentralized control versus centralized control- as the number of controllers increases. We present provably convergent algorithms for calculating Nash equilibrium prices for both the uncapacitated and capacitated cases and -using these algorithms- illustrate counterintuitive situations in which consumer surplus increases after decentralization. While we develop our model in the context of airline pricing, it is applicable to any service network such as freight transportation, pipelines, and toll roads as well as to the more general case of supply chain networks. In the rest of the dissertation, we focus on understanding and improving pricing decisions in the case when corporate headquarters set a list price for all products but local sales force is given discretion to adjust (or negotiate) prices for individual deals. This form of pricing is called list pricing with discretion (LPD) and it is commonly found in most business-to-business markets and in certain business-to-consumer settings, including consumer lending, insurance, and automobile sales. In the LPD setting, the question of how much (if any) pricing discretion should be granted to local sales force is crucial. In the second part of this thesis, we study this issue using two data sets - one from an online lender who sets all prices centrally and one from an indirect lender with local pricing discretion. We find strong evidence that the indirect sales force adjusts prices in a way that improves profitability. However, we also show that using a centralized, data-driven pricing optimization system has the potential of improving profitability further. In addition, using a control function approach, we show that the discretion applied by the local sales force introduces significant endogeneity into the indirect lender's pricing process. Ignoring this endogeneity can lead to severe underestimation of price sensitivity. These insights are valuable for any customized pricing market in which in-person interaction is part of the price-setting process. Finally, in the last part, we focus on the underlying negotiation process of the LPD setting and on the fact that not only buyers differ in their willingness-to-pay (WTP) but sellers also differ in the minimum prices (reservation prices) that they are willing to accept for the transaction. We develop a methodology based on the Expectation-Maximization (EM) algorithm to estimate both the WTP and the reservation price distributions given transactions data. The required data include information about both completed trades and failed trades, however price information is only available for completed trades (which is the most common situation in these markets). Using the same data from the auto lending industry, we show that our approach provides improved estimates of customer price-sensitivity over the approaches commonly used in practice. We also show how the WTP and reservation price estimates can be used to improve profits for the seller by optimally setting reservation prices on negotiations.



  • thumnail for Simsek_columbia_0054D_11638.pdf Simsek_columbia_0054D_11638.pdf binary/octet-stream 1.48 MB Download File

More About This Work

Academic Units
Thesis Advisors
Phillips, Robert
Ph.D., Columbia University
Published Here
October 17, 2013