A decision to buy an item at a regular price or wait for a possible markdown involves a multi-dimensional trade-off between the value of the item, the delay in getting it, the likelihood of getting it, and the magnitude of the price discount. Such trade-offs are prone to behavioral anomalies by which human decision makers deviate from the discounted expected utility model. We build an axiomatic preference model that accounts for three well-known anomalies, and produces a parsimonious generalization of discounted expected utility. We then plug this behavioral model into a Stackelberg-Nash game between a firm that decides the price discount and a continuum of consumers who decide to wait or buy, anticipating other consumers' decisions and the resultant likelihood of product availability. We solve the markdown management problem and contrast the results of our model with those under discounted expected utility. We analytically show that accounting for the behavioral anomalies can result in larger markdowns and higher revenues. Finally, we calibrate our model via a laboratory experiment, and validate its predictions out-of-sample.
Stored only when both papers are in the corpus — 43 citations exist upstream.