Research On Pricing Strategy Under Reference Price Effect

With the change of times and the development of economy,customers are becom-ing more and more "shrewd" during their purchasing process either online or offline.Especially in the retail industry,where information spreads extremely quickly,cus-tomers would generally use tools to track and observe the historical prices of the prod-uct,and form a price expectation(referred to as "reference prices”)for the product from these historical prices before making their purchasing decisions.Previous literatures have shown that reference price significantly affects consumers' purchasing behavior and thus has a great influence on retailers' pricing decisions.In addition,consumers usually have different risk attitudes towards the difference between the current sell-ing price and the reference price,which also significantly affects the retailers' pricing decisions.Meanwhile,with the development of network information technology,the retailers have gradually realized the multiformity and complexity of the customers' be-haviors and implemented many measures,such as demand learning and price-matching policy,to improve the total revenue.Therefore,it is increasingly necessary to explore the joint influences of the reference price effect and these measures on the retailers'pricing decisions.This dissertation explores the joint influences of the reference price effect and the factors that significantly affect the retailers' pricing decisions under different sales en-vironments,the research results provide certain guidance and referential significance for the retailers' pricing decisions.The main research contents and conclusions of this dissertation are as follows:In Chapter 3 of this dissertation,we consider a seller sells a single product over a finite horizon,with the objective of maximizing the expected total discounted revenue by dynamically adjusting posted prices.One distinct feature of our problem is that the customers' arrival rate is unknown to the seller but can be learned in a Bayesian method.Moreover,Chapter 3 assumes that the purchase behavior of the arriving consumer will be affected by the reference price.We formulate this problem as a Bayesian dynamic programming.First we analyze the structural properties of the optimal revenue function and the optimal pricing policy.We find that the problem can be substantially simplified in the case of sufficient inventory and demand learning can be decoupled from pricing decision.Then,we investigate the value of the market size(customers' arrival rate)and the effect of the reference price.Furthermore,we conduct several numerical examples to justify our theoretical results,examine the influence of demand learning,and we find that ignoring the reference price effect will lose substantial revenue,and bayesian demand learning may not improve revenue when inventory level is limited.In Chapter 4 of this dissertation,we consider a retailer that sells a product over a two-period horizon.The goal is to investigate the single and combined effects of the reference price and the price-matching policy on the purchasing behavior of consumers,and determine how these factors influence the retailer's optimal pricing decisions and optimal total discounted revenue.We first present a discrete-time dynamic pricing(DP)model over a two-period horizon with reference price effect in the presence of strategic consumers.Then we subsequently extend to another DP model in which the retailer implements a price-matching policy.The results show that under the reference price effect,the retailer's revenue will always decrease,even when a price-matching policy is implemented.The price-matching policy is not always beneficial for the retailer,especially when the discount factor is infinitely close to 1.Then,Chapter 4 proposes a"model misspecification" to investigate the effect of the reference price.Furthermore,we numerically explore the value of a price-matching policy.The results show that the value of a price-matching policy is generally the best when the discount factor is at a threshold value and the value is much greater in the presence of strategic purchasing behavior.In Chapter 5 of this dissertation,we consider a seller sells a single product in a short period and taking reference price effect into account in the presence of risk prefer-ence customers.Customers' demand for the product is closely related to their purchase probability,which is determined by their purchase utility that is contingent on the ref-erence price,selling price and risk coefficients,through a Multinomial Logit(MNL)model.Customers in the market are categorized as three types according to their asym-metry perceptions anchoring on the difference between the reference price and selling price:loss-averse,gain-seeking and loss-neutral.We first theoretically explore the in-fluences of the reference price,risk coefficients as well as the numbers of three types of customers on the seller's pricing decisions and profits.Then,we introduce model misspecification and use a computational study to further illustrate the significance of the seller's correct cognition on the customers' risk preference behaviors.We find that customers with higher reference price or lower risk coefficients would urge the seller to increase the optimal price.When there are more loss-averse customers in the mar-ket,markdown is optimal for the seller;whereas markup is optimal for the seller when more gain-seeking customers in the market.Besides,Chapter 5 shows that customers'risk preference behaviors are nonnegligible for the seller in making pricing decisions,especially when gain-seeking customers' risk coefficient is small enough or loss-averse customers' risk coefficient is large enough.The main contributions of this dissertation include:(1)Chapter 3 innovatively combines the consumer's reference price effect with the retailer's bayesian demand learning and analyzes the optimal pricing strategy in the framework of Bayesian dy-namic programming;(2)Chapter 4 is the first to combine the reference price effect and price-matching policy,and explores the joint influences of the reference price effect and price-matching policy on the retailer's pricing decisions for the strategic consumers;(3)Chapter 5 considers that there are three different types of consumers in the market.Different from the previous literature which assumed that the reference price directly affected the demand of product,Chapter 5 assumed that the reference price affected the purchase utility,which is transformed into the purchase probability through an MNL model.