Principles for effective automation in post-trade processing: Part 2 — thought experiments for friction-free trade processing

Abstract

This is the second of a two-part series. Part 1 published in an earlier issue of this journal described a model for gathering the data needed to understand how well automation was working in trade processing and where to focus investment on. Collecting the right type of data in a cost-effective way is essential to understanding which problems to focus on and also measuring the success of solutions put in place. Is it possible, however, to start with a clean slate where most sources of friction and complexity in trade processing simply cannot exist? This paper defines two models for trade processing that seek to eliminate the core sources of problems in three areas: trade capture, including trade enrichment; creating and maintaining consensus on the details of a trade; and correctly and consistently generating the trade events that come after execution, including those related to settlement. The first model, ‘T-Instant’, is based on creating a completely centralised and vertically integrated utility for trading, trade processing and issuance of securities. The second is a more distributed model called the ‘Continuous Consensus Model’ where parties to a trade can agree on a data model and data processing model that are combined with the ‘pairing’ rather than the matching of trades. This model allows for the continuous real-time identification of differences and also the sharing of data to help with a consistent view of a trade between all parties involved.