Progress report: Computer Aided Market Engineering
May 2003 - December 2003

Team

Project 4 began in May 2003 and is co-funded by the German Alexander von-Humboldt Foundation as a part of the TransCoop project. At this stage the project is mainly concerned with the conceptual foundations of market engineering in a broad sense and with the analysis of the process of engineering electronic negotiations.

Market engineering

Market engineering is pinpointed as an iterative approach ranging from the analysis of the social environment, to the extraction of formal requirements, the design of electronic negotiation protocols and to the implementation and evaluation of this process The conceptualization of the process is important for its computer-aided support. The activities focus on the foundations of the design and development of electronic negotiations as a core component of electronic markets. As electronic markets are being designed and evolving, it is important to understand how they and the underlying electronic negotiations work. The main goal of the market engineering project is to identify design patterns that will enhance the engineering process. In addition to practical approaches, design patterns also utilize tool support. It is expected that the use of tools will lead to greater efficiency and efficacy of the process of engineering electronic negotiations. The envisioned tool support comprises a pattern library, a decision support system for market patterns, a simulation and evaluation shell, and a code generator.

The project involves a number of graduate students who authored and co-authored journal publications, and made presentations at scientific and professional meetings.

Ph.D. and M.Sc. theses:

Al Hunaty, S. "Design Patterns Library for eNegotiations." M.Sc. thesis, (in progress).

Bendrich, T."Computational Mechanism Design." M.Sc. thesis, University of Karlsruhe, Karlsruhe, Germany (in progress).

Budimir, M. (2003). "Innovative Market Structures and Information Processing in Financial Markets." Ph.D. thesis, University of Karlsruhe, Karlsruhe, Germany, (in German).

Holtmann, C. "Organization of Electronic Markets - Market Engineering for Securities Trading." Ph.D. thesis, University of Karlsruhe, Karlsruhe, Germany (in progress).

Neumann, D. "Market Engineering as a Computer-aided Process." Ph.D. thesis, University of Karlsruhe, Karlsruhe, Germany (in progress).

Schnizler, B. "Iterative Combinatoral Auctions." M.Sc. thesis, University of Karlsruhe, Karlsruhe, Germany (in progress).

Strecker, S. “Electronic procurement with multiattribute auctions: A laboratory study”, Ph.D. thesis, University of Karlsruhe, Karlsruhe, Germany (in progress).

Strub, T. "Genetic Algorithms and Auction Evaluation." M.Sc. thesis, University of Karlsruhe, Karlsruhe, Germany (in progress).

Journal publications

Bichler, M. and J. Kalagnanam (2003). "Bidding Languages and Winner Determination in Multi-attribute Auctions." European Journal of Operational Research (forthcoming).

Bichler, M. and J. Kalagnanam (2004). "A Non-Parametric Estimator for Reserve Prices in Procurement Auctions." Information Technology Management Accounting Research (forthcoming).

Ehrhart , K.-M., C. Hoppe, J. Schleich and S. Seifert (2003). "Strategic Aspects of CO2-Emissions Trading: Theoretical Concepts and Empirical Findings." Energy and Environment (forthcoming).

Weinhardt, C., C. Holtmann and D. Neumann (2003). "Market Engineering." Wirtschaftsinformatik, (forthcoming, in German).

Seifert, S. and K. M. Ehrhart (2003). "Design of the 3G Spectrum Auctions in the UK and in Germany: An experimental analysis." German Economic Review (submitted).

Books and book chapters

Veit, D. J. (2003). Matchmaking in Electronic Markets. Springer-Verlag, (forthcoming).

Bichler, M. and J. Kalagnanam (2004). "Industrial Procurement Auctions. Combinatorial Auctions." in P. Cramton, Y. Shoham and R. Steinberg (eds.), (forthcoming).

Conference proceedings

Czernohous, C., K. Kolitz, J. Maekioe, I. Weber and C. Weinhardt (2003). "Integrating electronic market models - problems and solutions of parallel markets." 10th Research Symposium on Emerging Electronic Markets (RSEEM), Bremen.

Holtmann, C. and D. Neumann (2003). "Market and firm - two sides of a coin." 10th Research Symposium on Emerging Electronic Markets (RSEEM), Bremen.

Strecker, S. and S. Seifert (2003). “Preference Revelation in Multi-Attribute Bidding Procedures: An Experimental Analysis”, Proceedings of the Fourth e-Negotiations Workshop at the 14th International Conference on Database and Expert Systems Applications, DEXA 2003, , Prague, 850-854.

Strecker, S. (2003). "Preference revelation in multi-attribute reverse English auctions: A laboratory study." Proceedings of Twenty-Fourth International Conference on Information Systems (ICIS 2003) , Seattle, WA, USA, December 14-17, 2003, pp. 271-282..

Strecker, S. and S. Seifert (2004). "Electronic sourcing with multi-attribute auctions." Proceedings of the 37th Annual Hawaii International Conference on System Sciences (HICSS'04) , Big Island, Hawaii, January 5-8, 2004 , p. 70165b. .

Veit, D. J. and C. Czernohous (2003). "Automated bidding Strategy adaption using learning agents in many-to-many e-markets." Agent Mediated Electronic Commerce V (AMEC-V) , Melbourne.

Electronic negotiations

The foundations of electronic negotiations are extremely important, as they mainly constitute the domain theory that Market Engineering is exploiting. The first step was the derivation of a taxonomy that widely captures all essential aspects concerning negotiation protocols, negotiation situation and negotiation performance. With the conceptual elaboration of the Montreal Taxonomy existing approaches are extended such that a very broad variety of negotiation protocols are included. The Montreal Taxonomy was, furthermore, tested by confronting it with empirical negotiation systems. The evaluation yielded that the taxonomy is capable of sketching practical negotiation systems in terms of its concepts. For design purposes some amendments are necessary as the formulization those concepts.

The main results of the sub-project electronic negotiations are the derivation of a domain theory of electronic negotiations. A domain theory requires firstly a conceptualization of the domain and more importantly how these concepts are related. The conceptualisation is attacked with the refinement of the Montreal Taxonomy. In a subsequent step the development of models that explain links between those concepts is in the centre of attention.

The project involves a number of graduate students who authored and co-authored journal publications, and made presentations at scientific and professional meetings.

Ph.D. and M.Sc. theses:

Gimpel, H., "Auktionsverfahren im Aktienprimaermarkt – Experimentelle Evaluierung der Informationseffizienz." M.Sc. thesis, (in German).

Rolli, D. "Eine dynamische Workflow-Standard-Erweiterung im Hinblick auf die Steuerung elektronischer kombinierter Verhandlungen." M.Sc. thesis, (in German).

Journal publications

Bichler, M., G. E. Kersten and S. Strecker (2003). "Towards a Structured Design of Electronic Negotiations." Group Decision and Negotiation, 12(4): 311-335.

Bichler, M., G. E. Kersten and C. Weinhardt (2003). "Electronic Negotiations: Foundations, Systems and Experiments I." Group Decision and Negotiation, 12(2).

Bichler, M., G. E. Kersten and C. Weinhardt (2003). "Electronic Negotiations: Foundations, Systems and Experiments II." Group Decision and Negotiation, 12(4).

Neumann, D., M. Benyoucef, S. Bassil and J. Vachon (2003). "Applying the Montreal Taxonomy to State of the Art E-Negotiation Systems." Group Decision and Negotiation, 12(4): 287-310.

Stroebel, M. and C. Weinhardt (2003). "The Montreal Taxonomy for Electronic Negotiations." Group Decision and Negotiation, 12(2): 143-164.

Conference proceedings

Gimpel, H., H. Ludwig, A. Dan and B. Kearney (2003). PANDA: Specifying Policies for Automated Negotiations of Service Contracts. 1st International Conference on Service Oriented Computing (ICSOC 03), Trento, Italy.

Guo, Y., J. P. Mueller and C. Weinhardt (2003). Learning User Preferences for Multi-attribute Negotiation: An Evolutionary Approach. CEEMAS. V. Marík, J. P. Mueller and M. Pechoucek. Prague, Springer.

Plan 2004-2006

It is intended to extend the foundations activities in both areas. Moreover, based upon a solid fundament the development of CAME tools is envisioned to start in the first quarter of 2004. An overview of the planed activities for the 2004-2006 period is given in Figure 1.

Figure 1. Roadmap

Market engineering and electronic negotiations

Analysis of the social environment has up to now not received much of attention. However, this analysis is a major factor of Market Engineering as it extracts the requirements the design has to satisfy. If the requirements are fuzzy or even wrong, the design of negotiation protocols may not attain its goals.

One powerful tool to support this analysis task is the extension of the Montreal Taxonomy towards the negotiation situation. As the negotiation situation corresponds with the social environment the two pillars Market Engineering and Electronic Negotiations converge (Figure 1, P1)

Another important sub-project is the empirical evaluation of negotiation protocols. Currently, eBay renewed its protocol set by including a fixed price option to the existing auction. As it receives much of attention, the impact of such a protocol is evaluated by using laboratory experiments. (Figure 1, P2)

Computer aided market engineering (CAME)

The development and implementation of CAME tools will stand in the focus of the project. Three major tools are currently being planned:

1. A negotiation protocol configurator that is coupled with the market server. The configurator is intended to generate the code for the protocols automatically. The description language, the so-called generic market modelling language, has its roots in the Montreal Taxonomy (Figure 1, P3).
2. A design support tool. The design support tool is perceived to be a complement to the configurator, as it should semi-automatically or fully-automatically suggest an appropriate protocol (Figure 1, P4).
3. An experimental evaluation shell. This shell is assumed to support the evaluation phase by providing a tool-set for laboratory experiments as well as simulations (Figure 1, P5).

Cooperation

The CAME project is currently involved with the following companies:

• Reuters
  Reuters supplies data of live markets to the project in order to establish a knowledge base. Additionally, Reuters assists the conduct of real-time simulations on the market server concerning different market structures. This dedicated study allows Reuters to enhance their trading services by better addressing customers’ needs. The integration of the market server with affiliated Reuters products such as OrcTM is currently in consideration.
   
• Stuttgarter Boerse
  The Boerse Stuttgart – the European leading warrant exchange and Germany's most successful exchange for retail investors – is currently co-operating in the further refinement of the market engineering process. In several common organized projects the market engineering process is practically applied and in subsequent feedback-loops improved. Currently, two steps of market engineering process, namely the evaluation and re-design of their market structures and the introduction of innovative products are in centre of cooperation.
   
• Trading fair
  Trading fair is a spin-off enterprise of the chair for information management and systems. Their main focus is on the development and the operation of electronic market systems. Currently Trading Fair implements applications in order to facilitate multi-channel access to financial systems, in particularly electronic market places. The cooperation is founded on two pillars. Firstly, Trading Fair contributes its expertise in the field of market server implementations. Secondly, a multi-channel access to the electronic market is currently designed.

The cooperation with business partners shows that CAME is not just a concept; it is (partially) used in practice. In cooperation with the above three organizations, market engineering is practically in use in the area of financial markets. Recently, the domain of financial markets has been widened to the area of emission trading. This diversification may demonstrate the robustness of the developed approach and the tools that are subject to research within the SSHRC-project.

Planned Events

• Meeting along the DEXA 2004 conference in Zaragoza, Spain in August 2004.
• Joern Weigelt, MSc. Student, started his research stay in Montreal at the CIRANO in November 2003.