Agent-intermediated electronic markets in international freight transportation

https://doi.org/10.1016/j.dss.2004.10.008Get rights and content

Abstract

In many industries, agent-intermediated markets are inefficient because information about latent demand and supply never gets to market. We demonstrate how information technology (IT) in the form of an agent-intermediated electronic market (EM) alleviates this problem by enhancing the agent-as-market-maker using the international freight transportation industry as an example. We find that an EM increases agent participation and investment thereby increasing demand and supply. Because of tradeoffs between incentives for investment, the EM chooses a profit allocation between agents resulting in limited agent participation. In addition, when price depends on demand and supply balances, price and volume in the market can increase simultaneously.

Introduction

In many industries, market institutions are inefficient because information about latent demand and supply never gets to market and many feasible transactions between buyers and sellers are not completed. Characteristics of such markets include insufficient numbers of buyers and sellers (thin markets), heterogeneous and complex goods (specificity), and information asymmetries which may hold-up trade. When only partial information about demand and supply is available, individual demand and supply are not properly matched, and markets do not always clear. Consequently, incentives to bring new demand and supply information to the market are reduced. The “traditional market” in many industries uses an agent as an intermediary between customer and supplier to overcome these causes of inefficiency.

Two new institutions based on electronic markets (EMs) have recently emerged. The first substitutes an EM for the agent in the traditional market. This substitution occurs in industries where the quality of the goods can be assessed at low cost, and the transfer of ownership and possession is straightforward. The second EM supports trade between agents. This latter case, which we call an agent-intermediated EM, is the focus of our paper. This support is necessary in industries where the quality of goods is hard to measure and/or the transfer of ownership and possession is complex. In these industries specific domain knowledge held by agents is vital to quality assurance and post-transaction market clearing operations. This view is consistent with the trade press that suggests the critical lesson learnt in electronic marketplaces so far is to specialize [13]—that is, understanding the special characteristics of a given marketplace is crucial for performance. It is also consistent with transaction cost economics, which posits that because of administrative costs, distortions, and inefficiencies from low-powered incentives, most markets have evolved to keep transaction risk low and preserve specialized information the market needs to function efficiently [27], [28].

Some believe that information technology (IT) through its cost-effective enabling of EMs will be used to disintermediate markets [11], [15], [26]. However, in the most commonly used examples—on-line securities trading and air travel reservations—markets have not been disintermediated. Rather, the intermediary has been (partially) automated. On-line trading requires a broker to carry out the transaction and most on-line airlines reservations are done through travel agents that have an Internet presence. Indeed, many believe that there will not be disintermediation, just new types of intermediaries [3]. A classic example is Aucnet—the electronic auction for used cars in Japan. Aucnet is an agent-intermediated EM with dealers acting as agents for their customers in the buying and selling of cars. Supporting the value of agent intermediation, research has found that the average contract price of secondhand cars sold through Aucnet is much higher than that of traditional, non-electronic markets [16]. New types of intermediaries provide information to buyers and sellers, aggregate demand or supply, manage physical deliveries and payments, and provide trust and integrity to the EM [4]. Intermediaries such as trusted third parties have been proposed as a way of providing quality assurance [2]. Trust is critical to quality assurance, and it has been found to be an important determinant of EM adoption [7]. Even in financial markets where quality assurance of the process rather than the traded good is the issue, specialist auctions were found to increase market quality by reducing bid–ask spreads, transaction costs, and time to execution as compared to automated order matching systems [8].

IT has also been used to explain restrictions in the number of suppliers that participate in buyer–supplier relationships. For example, incentives can motivate a buyer to limit the number of suppliers it uses [5]. That is, to induce supplier investments in IT-related non-contractible assets, commitments by buyers are more credible if they limit the number of participating suppliers. The number of contract terms in a contract fall with increasing levels of IT due to more economical monitoring [6]. If suppliers face relationship-specific costs, the increases in IT can cause the buyer to limit its number of suppliers. Because of the negative externalities (e.g., competition) between suppliers, participation growth in interorganizational systems stalls without buyer subsidies [21]. Finally, a supplier's adoption of electronic data interchange (EDI) can generate positive externalities for buyers while producing negative externalities between suppliers [25]. Thus, it may be in the buyers' interest to subsidize suppliers' adoption of EDI. Regardless, if supplier adoption costs are high, then partial adoption by the supplier base may be optimal for the buyer.

An important example where demand and supply information in the market is incomplete is the less-than-full-load ocean shipping industry. Lumpy supply—a container—together with less-than-container size shipments often results in partially empty containers being shipped while partial container shipments that could have filled the partially empty containers remain on the dock. The information required to complete transactions in ocean shipping is extensive. Individual shipments have a specific origin and destination—possibly with a preferred routing, and restrictions on when they must depart and arrive. Available container capacity may not be the right kind of capacity because of shipment dimensions or special shipment requirements such as refrigeration. This market coordination problem is an instance of the more general less-than-full-load problem that prevails in all modes of the transportation industry. A recent report indicates that of a US$921B logistics market, 18% of capacity moves empty, leading to a collective loss of US$165B [18].

Agents, such as international freight forwarders, brokers, non-vessel operation common carriers (NVOCCs), and third party logistics providers (3PLs), serve as market-makers, providing search, price discovery, and numerous market clearing operations for both shippers and carriers. On the demand side, they provide shippers with financial services (letters of credit) and insurance, often assuming some of the liability of carriage. They also have expertise in managing the loading/unloading and pick-up/delivery process—in particular, export permits and customs clearance for international shipments [23]. Consequently, on the demand side, these agents need to invest in specialized knowledge such as customs regulations in particular regions in order to support the above activities. In addition, they need to invest usual demand-generating activities such as promotional visits to customers (i.e., shippers). On the supply side, they provide carriers with consolidation of shipments and the discovery of available less-than-full-load capacity. Agent investments on the supply side are investments in specialized knowledge of individual carrier operations such as methods used to consolidate freight, and contacts in ports and staging areas.

Agents are specialized to serve niche markets, for example, transporting individual automobiles from Europe to the US. As a result, the ocean shipping market has many agents because specialized knowledge resides with individual agents, and each agent's expertise differs from another's. Moreover, the knowledge on which agent expertise is based changes continuously [14]. Niche markets and specialized knowledge is the differentiation that mitigates head-on competition between agents. An example of an agent-intermediated EM in ocean shipping is GT Nexus (recently renamed from Tradient) which serves merchant importers and exporters, freight forwarders, 3PLs, and carriers. Testifying to the complexity of commercial shipping, and in our view the need for agent intermediation, “…a typical commercial shipment involves 9 different participants, 20 separate documents, 35 customer–vendor interactions, and 4 modes of transport. It all can take place over the course of several weeks or months, and can cross multiple international borders.” [12].

Underutilization of capacity occurs precisely because the traditional market is decentralized-individual agents try to match their own demand to their own supply, combining two aspects of each agent's activities—promotion to shippers (generation of demand) and provision of capacity (generation of supply). This coupling is evident when an agent purchases all or part of a container from a carrier and promotes that capacity to shippers or, alternatively, arranges for a shipment and then searches for the needed capacity. The problem is that agents frequently generate demand which they cannot individually satisfy with their generation of supply, and vice versa. Nonetheless, this traditional market still operates more efficiently than direct shipper (customer) to carrier arrangements because direct shipper to carrier arrangements cannot take advantage of the more extensive expertise an agent provides.

The agent-intermediated EM we propose broadens the traditional market by matching unsatisfied demand generated by one agent to underutilized supply provided by another, freeing agents from needing to match their own demand to their own supply. The EM allows agents to anonymously exchange information about capacity they need or have without revealing the identity of their shipper or carrier (prior to the transaction). It does not require that agents reveal all the demand and supply they generate—when an agent can match its own demand and supply it may operate as it always has in the traditional market. Nor does it preclude off-EM arrangements between individual agents. In these markets, agents invest in finding demand (shipments) or supply (capacity). Through the EM agents can be rewarded for generating either side of the transaction. Indeed, the agent-intermediated EM together with the traditional market can result in higher demand than a traditional market alone: Because of the difficulties and additional costs of finding capacity for idiosyncratic shipments, many shippers in traditional markets do not complete transactions they otherwise would if access to their specific transportation needs was less expensive.

In our example of ocean shipping, there have been two recent regulatory changes. Prior to 1984 ocean shipping was highly regulated and contract terms were standardized and public. The (US) Shipping Act of 1984 deregulated contracts to allow greater flexibility in the types of tariffs and contracts [17], [22]. In addition, the Act began the movement towards greater confidentiality such that only the essential terms of a contract (origin and destination) were published. Specifically, prices, capacity, and other contract details are now confidential. The Ocean Shipping Reform Act of 1998 went even further than the earlier Act by removing from public view origins and destinations for through intermodal movements, rates, liquidated damages for non-performance, and service commitments such as assured space, transit time, and port rotation. These regulatory changes have encouraged specialized contracts and terms, thereby fostering traditional markets and EMs we propose. Because traditional markets involve a single agent, these markets provide for the confidentiality of contracts and terms. The EMs we propose use IT specifically designed to protect the confidentiality and anonymity of their participating agents and their clients.

Our research objectives are to examine agent investment and agent participation in agent-intermediated EMs together with traditional markets as compared to investment and participation in traditional markets alone, and to examine the potential effects resulting from the addition of an agent-intermediated EM on market prices. Our main results are as follows. With an agent-intermediated EM, agent investment is higher and more agents participate, increasing demand and supply relative to the traditional market alone. However, the optimal allocation of profits between agents results in limited market participation where only a select group of agents participate. In addition, when price depends on demand and supply balances in the agent-intermediated EM—balances that are affected by the EM through increased agent investment—we find that both price and total volume may increase simultaneously. The limited market participation results because of tradeoffs between incentives for investment in demand and in supply: An increase in the allocation of profit for a given transaction to agents that generate demand raises demand generating investment and lowers supply generating investment (and vice versa). Simultaneous increases in price and volume can occur when an increase in price results in greater volume impacts through supply generating investment and agent participation than through demand generating investment and the direct effect of price on volume.

Apart from ocean shipping, other transportation industries use variants of agent-intermediated EMs. In the motor carrier transport, DAT Services [9] offers electronic freight matching to agents and carriers—providing electronic support to agents generating demand, but bypassing agent provision of supply. Indeed, carriers may carry out the agent function in generating supply in our model. Thus, our results generalize to agent-intermediated EMs where carriers are agents on the supply side. In addition, the National Transportation Exchange (NTE) [20] allows shippers or agents to post loads and member carriers to select loads to fill their unused capacity. NTE pre-authorizes shippers/agents and carriers, and serves standardized loads and domestic routes, thereby eliminating problems of quality assurance and ownership/possession. In a similar industry, air cargo, efforts by carriers to link directly with customers, thereby bypassing agents, failed because carriers did not account for the agents' special knowledge about customers and other carriers [10], [1].

The scenario we model is more broadly applicable than transportation. An example of an agent-intermediated EM outside the freight transportation industry is the Multiple Listing Service (MLS) in the real estate industry, a market that evolved in manual form prior to the Internet. That industry is dominated by agent intermediation rather than direct buyer-seller transactions, and has agents representing each side of a transaction—one that serves the buyer and another that lists property for the seller. Still other industries use agent-intermediated EMs. Headhunters use online job markets to match their job candidates to positions and their positions to candidates, brokerage houses use private electronic markets such as Liquidnet to adjust their portfolios, and travel agents use web-based markets for regular and charter services to satisfy customers.

The paper is organized as follows. We first explain the structure of our setting, the notation, and assumptions. Next we compare the traditional market to the agent-intermediated EM when prices are determined outside the EM. Then we consider when prices are affected by demand and supply balances in the EM, and explore the effect of the EM on welfare.

Section snippets

Structure of the setting

We represent agents by v∈[v, v¯], where a larger v corresponds to an agent with greater potential to generate demand and supply. An agent can have greater potential if it has a larger market niche it can serve with its expertise, or have more individual contacts with shippers or carriers—either would yield a greater return to its investment. v has the uniform density h(v)>0 over the support and is zero elsewhere. Agents decide whether to participate in the EM and how much to invest. There are

The traditional market

Each agent chooses its level of investment in generating demand and generating supply, dv and sv respectively. Throughout, we model agents as price-takers whereby each agent is not large enough for its individual investment decision to impact price. We examine the impact of a price determined by aggregate demand and supply conditions later. Agent profits are price times own demand, less costs of investment and less the cost of operations K,π(v,dv,sv)=pqa(v,dv,sv,p)dvsvK.

The first-order

Agent participation

We define the smallest agent that participates in the EM as =min{v|ψ(v, d, s, z)=0}. This marginal agent must satisfy the indifference conditionpqa(vˆ,dvˆ(p,f,m),svˆ(p,m),p)+[pfm]qb(vˆ,dvˆ(p,f,m),s\vˆ(p,m),z(vˆ),p)+mqc(vˆ,d\vˆ(p,f,m),svˆ(p,m),z(vˆ),p)dvˆ(p,f,m)svˆ(p,m)K=0=μ(p,f,m,vˆ),where μ(p, f, m, ) is an implicit function defining . Taking the partial derivative with respect to , we can use the first-order conditions in Eq. (2) to eliminate the indirect effects of through

Impact on the price

So far, we have considered the price as fixed by conditions outside of the EM such that both individual agents and the industry are price-takers. Although agents may not be able to influence the price individually, their aggregate investments in generating demand and supply may impact the price because they may change the balance between demand and supply. Endogenizing price means incorporating the difference between investments in generating demand and investments in generating supply.

Let the

Conclusion

We have shown that with price determined by the general freight transportation marketplace, an agent-intermediated EM increases agent participation and investment. The importance of agent investment is highlighted by the EM's allocation of profits between demand and supply providing agents on a given transaction whereby the EM balances participation and investment effects in its allocation, consequently restricting agent participation in the market.

This restriction of participation is different

Acknowledgements

Helpful suggestions have been provided by the Guest Editor and two anonymous Reviewers, seminar participants at Penn State University, Michigan State University, The Ohio State University, University of Southern California, University of Wisconsin-Madison, University of Oregon, University of Pittsburgh, and participants at WISE. We especially thank Clive Wrigley and John King for discussions about freight transportation. We also thank the National Science Foundation, the Natural Science and

Al Dexter is Professor Emeritus at the Sauder School of Business, The University of British Columbia, Canada. Before arriving at UBC he completed his MBA at Harvard Business School and Ph.D at Columbia University. His current research includes examining issues in information technology (IT) and administration of the IT function, economics of IT in new organizational forms, and more recently the impact of IT on broader economic measures such as inflation and productivity. His first publication

References (28)

  • S. Ba et al.

    Building trust in online auction market through an economic incentive mechanism

    Decision Support Systems

    (2003)
  • R. Gellman

    Disintermediation and the internet

    Government Information Quarterly

    (1996)
  • R. Alt et al.

    The great reversal: information and transportation infrastructure in the intermodal vision

  • J.P. Bailey et al.

    An exploratory study of the emerging role of electronic Intermediaries

    International Journal of Electronic Commerce

    (1997 (Spring))
  • J.Y. Bakos

    The emerging role of electronic marketplaces on the internet

    Communications of the ACM

    (1998 (August))
  • J.Y. Bakos et al.

    From vendors to partners: information technology and incomplete contracts in buyer–supplier relationships

    Journal of Organizational Computing

    (1993 (December))
  • R.D. Banker et al.

    Information technology, contract completeness, and buyer–supplier relationships

  • A.M., Chircu, G.B., Davis, R.J., Kauffman, The Role of Trust and Expertise in the Adoption of Electronic Commerce...
  • E.K. Clemons et al.

    Alternative securities trading systems: tests and regulatory implications of the adoption of technology

    Information Systems Research

    (1996 (June))
  • DAT Services, http://www.dat.com...
  • P. Forster et al.

    Information infrastructure standards in heterogeneous sectors: lessons from the worldwide air cargo community

  • GT Nexus, http://www.tradient.com/about...
  • M. Halper

    Meet the new middlemen

    Computerworld

    (1997 (April 28))
  • F.A. Hayek

    The use of knowledge in society

    American Economic Review

    (1945 (September))
  • Cited by (0)

    Al Dexter is Professor Emeritus at the Sauder School of Business, The University of British Columbia, Canada. Before arriving at UBC he completed his MBA at Harvard Business School and Ph.D at Columbia University. His current research includes examining issues in information technology (IT) and administration of the IT function, economics of IT in new organizational forms, and more recently the impact of IT on broader economic measures such as inflation and productivity. His first publication dates from 1966 and he has published well over a hundred articles in learned and professional journals including: The Accounting Review, Communications of The ACM, INFOR, Information Systems Research, IEEE Transactions on Engineering Management, Journal of Accounting Research, Journal of Money, Credit and Banking, Journal of Monetary Economics, Management Science and Marketing Science. He has consulted to numerous legal firms as an Expert Witness on matters of intellectual property and has given opinions relating to copyright infringement in The Supreme Courtj of British Columbia. He currently has grants from the major Canadian research institutions – NSERC and SSHRC.

    Dr. Barrie R. Nault is The David B. Robson Professor in Management (MIS) in the Haskayne School of Business at the University of Calgary. Before moving to Calgary he was on faculty at the Ohio State University, the University of California, and the University of Alberta. He received his Ph.D. from the University of British Columbia. His current research includes how information technology affects ownership, incentives, membership and investement in new organizational forms such as alliances, network and virtual organizations; the impact of information technology on economic indicators such as inflation and productivity; and the diffusion of new technology, examining when new technologies should be released and what incentives can be put in place to make them successful. Dr. Nault has published articles in numerous journals including Information Systems Research; IEEE Transactions on Engineering Management; Journal of Money, Credit, and Banking; Management Information Systems Quarterly; Management Science; Strategic Management Journal; Marketing Science; Journal of Monetary Economics, and Organization Science. He has written reports for the National Research Council, and has held grants from the National Science Foundation in the US as well as NSERC and SSHRC in Canada. He is a Departmental Editor for Management Science, and is an Associate Editor for Information Systems Research. He also serves on the editorial boards of the IEEE Transactions on Engineering Management, and Production and Operations Management.

    View full text