Market Design provides the structure that makes a market function better. The design of auction mechanisms, auction rules *, is at the core of market design. The application of auctions has a very long history but the theory of auctions is rather new and has developed along with the discipline of information economics *. An auction may improve the allocation of goods and services, e.g. by introducing a pricing mechanism that leads to more profitable trading, by concentrating the market, or by making the market more transparent.

Private information

The central difficulty in designing auctions is that of private (confidential and decentralized) information. The bidders have private information, e.g. about their preferences or production costs. This information is needed - directly or indirectly - to determine an optimal allocation.

On the other hand, economic agents cannot be trusted to reveal their private information unless they are given the right incentives to do so. There are at least two reasons why direct or indirect revelation cannot be accomplished for free. First, the agents may try to act strategically to influence the auction outcome in their own favor. By taking advantage of their information they might possibly get a better share of the reallocation gains. Secondly, even if strategic manipulations do not pay in the auction, e.g. because each agent is sufficiently small to have any significant impact, the agents may still be reluctant to reveal their private information since they may fear that it can be misused in later or parallel markets. 

Farmers, for example, may not trust a monopsonist * processor to run an auction to reallocate production contracts since they may fear that the processor will be able to learn the marginal costs of the individual farmers and use this in subsequent negotiations. For more see Cases >>

Direct revelation mechanisms

An important group of auctions provide incentives to reveal truthful information. They are typically called direct revelation mechanisms. A central result in mechanism design states that for any mechanism there exists a direct revelation mechanism to which the participants tell the truth, that yields the same result.

The idea of this so-called revelation principle is that an impartial mediator or social planner uses the revealed information in the best interest of the parties. To avoid that the parties would be better off giving false information, the planner must restrict the way he uses the information - and he must be able to commit to this restricted usage.

In many cases, this is neither trivially nor cheaply accomplished. A mediator may be tempted or bribed to misuse the information he acquires in pursuit of his own or a specific bidder’s particular interests. The lack of a trusted impartial “planner” - or a trusted third party - that can compute the optimal outcome, is hereby a major obstacle in practical mechanism design. Solving this central issue is a primary concern of the Partisia Market Design solutions. For more see Solutions >>

Economic context

The actual choice of market design depends on the particular economic context. In the context of private information, the task is to set up rules that provide the market participants with proper incentives to allocate the goods or services in the most efficient way in the particular context. Hereby the focus is on maximizing the private and public gains from trade.

In practice, the choice of sophisticated market design will be bounded by the involved cost of implementing, using and administrating the market. To summarize, the choice of market design must solve three problems:

1. Ensure that the market provides enough structure to solve the allocation problem
2. Ensure that the market participants have the right incentives to solve the allocation problem
3. Ensure minimal administrative costs and costs of using the market for the participant

The starting point of Partisia Market Design is to provide market designs (in particular auctions and exchanges) that best meet these objectives. Our primary product is software that implements and administrates such auctions at any desired level of confidentiality.

For more see Solutions >>

References

• Klemperer, P. 2004. Auctions: Theory and Practice, Princeton University Press.
• Milgrom, P. 2004. Putting Auction Theory to Work, Cambridge University Press.
• Myerson, R. 1979. Incentive-compatibility and The Bargaining Problem. Econometrica 47, 61–73.