First-ever large-scale economic application of multi-party computation

Earlier this year, three PCs in a network determined the clearing price in a double auction with some 2400 bids without ever decrypting these. The final price and the list of buyer and sellers were then decrypted. This represents the first ever large-scale application of so-called multi-party computation, and international experts consider this to be a major break-trough.

The actual application was also novel since it involved the first reallocation of production rights for sugar beets in Denmark by the establishment of a real market for such right. The application was developed in part in collaboration with Danisco.

From CBS Professor Peter Bogetoft participates and from the point of view of economics and the design of mechanism and institutions, the perspectives are promising. The technology developed can implement many alternative mechanisms that would normally require the introduction of a so-called trusted-third-party, e.g. an auction house, without the actual usage of such a party. It suffices that each of the parties trust a subset of the participants – possibly only themselves! In this way we can – in principle - implement any so-called direct revelation mechanism at very low costs, explains Peter Bogetoft.

The implementation is part of a major research effort, SIMAP, supported by the Danish Strategic Research Council. It involves the collaboration of computer scientist specialized in cryptology and economists specialized in mechanism design.

For further information, please see the SIMAP homepage, or contact Professor Peter Bogetoft, pb.eco@cbs.dk