Privacy Enhancing Technologies (P.E.T), or how to improve overall efficiency and reduce management costs for many use cases in the financial sector

Maintaining data quality is a real challenge for financial institutions. Beyond the costs involved, data compliance and updating are tasks that require considerable human resources.
Every year, banks purchase data worth hundreds of millions of euros from Data Vendors (firms that specialize in obtaining and reselling customer information for other companies). Data changes and needs to be updated both for business development purposes and to meet the demands of regulators, who pressure financial institutions to maintain their data at the highest possible quality levels.
Since even the Data Vendors are finding it difficult to preserve the quality of their own data, this is a big ask.

This is why, at the beginning of 2018, Societe Generale (via its Global Business Service Unit) joined a consortium called  MADRec (Mass Anonymous Data Reconciliation), to launch a platform used to measure the Group's data quality by comparing it with its peers and  Data Vendors. Several banks and Data Providers have progressively joined the consortium, attracted by the stakes of the project. Due to a different strategic vision, Societe Generale decided to withdraw from MADRec in September 2019, to set up its own consortium and create, on the same principle as MADRec, a new platform to assess whether P.E.T could enable the creation of a shared solution to :

• measure the quality of client data against its peers without revealing data and client relationships;

• avoid a costly third party to centralize checks; 

• reduce internal maintenance costs.

Thanks to the P.E.T, at the beginning of 2020 we’ve been able to reach the right level of trust among us and run proof of concepts that validate these assumptions. This is the DANIE project.

The DANIE project brings together a consortium of banks and data providers collaborating to develop an innovative industry solution that will help improve data quality, reduce costs and drive standardization for critical reference data required for trade processing, risk management and regulatory reporting.
The consortium’s objective is to improve data quality by using a distributed reconciliation capability (P.E.T) to securely and anonymously highlight each firm’s own data anomalies against equivalent values submitted by peers. The technical solution is provided by a London based startup called Secretarium.
DANIE’s consortium started with our first use case to benchmark reference data quality on third parties with our peers via a matching key: the Legal Entity Identifier (LEI). All data fields linked to this LEI could then be reconciled with our peers.
Each participant sent his list of LEI to DANIE’s platform, anonymously and encrypted, with respect to the format of each data field required by the platform. 
Depending on how large the dataset is, it takes from seconds to a few minutes to get a report showing if you are in full consensus with your peers and where you have anomalies.
The major difference compared to the current approach, is that we can compare our data directly with our peers respecting privacy and without a trusted third party. 
Three banks and three data providers participated in the POCs. The first one took place in March 2020 with 22,570 LEI on 15 data fields. The second one was done in June: DANIE’s consortium extended the volume of LEI to 200K with around 30 data fields. In a matter of seconds, banks were able to identify 5 to 10% of errors among thousands of data items, using a user-friendly platform.

The consortium is planning to extend the data set to Anti-Money Laundering (AML) /Know Your Customer (KYC) areas. The DANIE Consortium continues to onboard new Financial Institutions as this solution can only be valuable and efficient if many players/banks join the exercise. Thanks to the scalability of the confidential computing provided by Secretarium, we need to think of DANIE as a platform where we can add several applications. A data quality Benchmark is the first application, but we can think of applications for Anti-Money Laundering, Counter Financing Terrorism regulation, Market Data etc., as this new technology paves the way to many opportunities.