Introduction

Abstract

Today’s transportation market is far from perfect. Market participants have admission only to a certain fraction of the market and price formation lacks transparency. However, this state is no law of nature since in principle, all market participants could manage to get together somehow and do business — but gaining new business partners beyond existing bonds usually takes too high efforts. After all, the items in transportation markets are usually traded under extreme time pressure: transportation orders have fixed pick-up and delivery deadlines and free transportation capacity cannot be used thereafter. Here, Internet-based freight markets promise help. Since the early 1990s many of such markets have come into existence so that a considerable number exists today. For Germany alone, over 50 different electronic transportation marketplaces are listed in a database hosted at the University at Bremen (2001). The mediation models of these marketplaces are quite diverse (cp. Ihde (2002), Schneider et al. (2002)). Electronic blackboards with or without fixed fees, open only to a certain part of the market, can be found as well as auction houses and real exchanges with transaction-based fees (Teleroute (2001), Benelog (2001a), Eulox (2001a)). Nevertheless, one important aspect has been ignored up until now: carriers usually have non-additive preferences with respect to combinations of transportation orders. The underlying reason is that capacity utilization of the respective vehicles can be maximized by forming round trips, which will generally bring down transportation costs significantly. In the simplest (and ideal) case, a carrier who is located in A has one order from A to a location B and one order from B to A that he can fulfil both sequentially with one and the same vehicle. If two such ‘complementary’ orders are auctioned off packagewize, both sides of the market will benefit — carriers, because they have a maximal capacity utilization granted and can make better prices without shrinking margin. Shippers benefit from these better prices. Auctioning off two complementary orders as a package is easy as long as these orders belong to the same shipper. The crux, however, is that in transportation spot markets they will generally stem from two different shippers. This has two consequences. First, two shippers with complementary orders do have to know about this and they do have to get into contact with each other. Second, the hammer price for an order package will be a package price and the involved shippers have to agree about their respective shares. In a highly dynamic business like the transportation business with little time for coordination, these are serious obstacles. However, both obstacles could be overcome. The first problem can be tackled with software routines that automatically identify complementary orders. Such a routine has been developed at DaimlerChrysler’s research department. The second problem requires the design of a mechanism that allows for a packagewize placement of orders from two different shippers. Obviously, the mechanism involves the collection of orders, the aggregation of complementary orders to order packages, the placement of orders and order packages, and the division of package prices. This second issue is what this thesis is about.