Optimal pricing to minimize maximum regret with limited demand information

Ming Chen,Zhi-Long Chen

DOI: https://doi.org/10.1016/j.cor.2020.105070

2020-12-01

Abstract:<p>In this paper, we consider a pricing problem faced by a seller that sells a given inventory of some product over a short selling horizon with limited demand information. The seller knows only that the demand is a linear function of the price, but does not know the parameters involved in the demand function. However, the seller knows that each parameter involved in the demand function belongs to a known interval. The seller's objective is to determine the optimal price for the entire selling season to minimize the maximum regret, where the maximum regret is defined as the maximum possible loss of revenue due to not knowing the precise values of the parameters. We derive closed-form optimal solutions for the problem under all possible cases of input parameters and identify some structural properties of the solution. We conduct computational tests to compare our modeling approach with several benchmark approaches and report related insights.</p>

computer science, interdisciplinary applications,engineering, industrial,operations research & management science

What problem does this paper attempt to address?

### The problems the paper attempts to solve

This paper explores the problem of how a vendor can determine the optimal price to minimize the maximum regret under limited demand information. Specifically:

1. **Background and challenges**:
- The vendor sells a certain inventory of products within a relatively short sales season.
- The vendor only knows that the demand is a linear function of the price \( D(P)=\max(a - bP,0) \), but does not know the specific values of the parameters \( a \) and \( b \).
- The parameters \( a \) and \( b \) are restricted within known intervals, that is, \( a\in [a_1,a_2] \) and \( b\in [b_1,b_2] \).
- The goal is to determine the optimal price throughout the sales season to minimize the maximum possible revenue loss (maximum regret) due to the unknown parameters.

2. **Problem definition**:
- **Maximum regret**: Defined as the difference between the total revenue of the ideal pricing scheme (i.e., knowing the exact value of future demand) and the total revenue of the actually used pricing scheme.
- **Objective**: Find a price \( P \) such that the maximum regret is minimized under all possible combinations of demand parameters.

3. **Research methods**:
- Use the **minimax regret** criterion to optimize price decisions.
- Derive a closed - form optimal solution by analyzing the optimal solutions in different situations.
- Conduct computational experiments to compare the performance of the proposed model with several benchmark models.

4. **Main contributions**:
- Derive the closed - form optimal solutions to the problem in various situations.
- Prove that the maximum regret is a quasi - convex function and that the optimal price is unique under all combinations of input parameters.
- Through computational experiments, demonstrate the superiority of the proposed method in reducing the maximum regret and increasing the average revenue.

### Summary

The paper aims to solve the problem of how to minimize the maximum regret by determining the optimal price under limited demand information. By analyzing the optimal solutions in different situations, the author provides an effective pricing strategy that can achieve higher revenue stability in an uncertain market environment.