Dynamic Price Forecasting in Simultaneous Online A

Authors: Mayukh Dass Wolfgang Jank Galit Shmueli

Publish Date: 2010

Volume: , Issue: , Pages: 417-445

PDF Link

Abstract

In recent years the simultaneous online auction SOA has become a popular mechanism for selling heterogeneous items such as antiques art furniture and collectibles These auctions sell multiple items concurrently to a selected group of bidders who often participate in multiple auctions simultaneously Such bidder behavior creates a unique competitive environment where bidders compete against each other both within the same auction as well as across different auctions In this chapter we present a novel dynamic forecasting approach for predicting price in ongoing SOAs Our proposed model generates a price forecast from the time of prediction until auction close It updates its forecasts in realtime as the auction progresses based on newly arriving information price dynamics and competition intensity Applying this method to a dataset of contemporary Indian art SOAs we find high predictive accuracy of the dynamic model in comparison to more traditional approaches We further investigate the source of the predictive power of our model and find that price dynamics capture bidder competition within and across auctions The importance of this finding is both conceptual and practical price dynamics are simple to compute at high accuracy as they require information only from the focal auction and are therefore a parsimonious representation of different forms of withinauction and betweenauction competition

Keywords: