Initial Public Offering (IPO) on Permissioned Blockchain using Secure Multiparty Computation


Authors: Fabrice Benhamouda (IBM), Angelo DeCaro (IBM), Shai Halevi (IBM), Tzipora Halevi (Brooklyn College), Charanjit Jutla (IBM), Yacov Manevich (IBM), and Qi Zhang (IBM)

Abstract: In this work, we design secure multiparty computation capabili- ties into the permissioned blockchain architecture of Hyperledger Fabric and use them to implement the clearing price mechanism for IPOs. Fabric is a platform for "blockchain for business," providing a robust mechanism for a set of "peers" to jointly maintain a distributed immutable ledger. As with any blockchain, using confi- dential data on the ledger is a challenge, and we use cryptographic secure multiparty computation (MPC) to address this challenge. The use of secure MPC to enable confidential data on blockchains was proposed before, but this approach only now begins to find its way to real-life systems. In particular, ours is the first attempt to integrate it in a complex blockchain environment such as Fabric.

To enable this approach, we integrated mechanisms into Fabric to let the peers access local information such as their respective keys, and also send messages to each other while executing smart contracts. We added to Fabric a library that implements the required cryptographic tools, and made it accessible from the smart contract. We then designed an efficient cryptographic protocol for the IPO clearing price mechanism, and used our integrated system to run it on Fabric. Although not fully optimized yet, the performance of the resulting implementation is more than fast enough for this particular application, ranging from 8 to 23 seconds to execute an IPO sale.

Keywords. Blockchain, Confidentiality, Hyperledger Fabric, Integration, IPO, Secure Multiparty Computation

Availability: Paper available as PDF, 1.2 Mbyte.


Shai Halevi's home page.