Dynamic pricing and inventory management with demand learning: A bayesian approach.

Area of Science:

• Operations Research
• Supply Chain Management
• Econometrics

Background:

• Retail firms face volatile markets with price-sensitive, random demand of unknown distribution.
• Dynamic inventory replenishment and pricing are crucial for managing durable goods in such environments.

Purpose of the Study:

• To develop an optimal inventory and pricing policy for a retail firm selling durable products under demand uncertainty.
• To incorporate demand learning into inventory and pricing decisions using an empirical Bayesian approach.

Main Methods:

• Formulated the problem as a stochastic dynamic program using an empirical Bayesian approach.
• Identified regularity conditions for demand models to establish an optimal policy.
• Employed dimensionality reduction for a more tractable normalized dynamic program.

Main Results:

• The state-dependent base-stock list-price policy is proven optimal under specific demand model conditions.
• Demand learning significantly impacts the optimal policy compared to a system without Bayesian updates.
• Analysis extended to scenarios with unobserved lost sales and additive demand.

Conclusions:

• The empirical Bayesian approach provides a robust framework for optimizing dynamic pricing and inventory for durable goods with uncertain demand.
• Demand learning is a critical factor in enhancing the efficiency of inventory and pricing strategies in volatile retail markets.