2014年-FSB全球金融稳定委员会_Feasibility_Study_on_Approaches_to_Aggregate_OTC_Derivatives_Data_95页_1mb

Summary of the Feasibility Study on Approaches to Aggregate OTC Derivatives Data

Core Content

This feasibility study, published on 19 September 2014, was commissioned by the Financial Stability Board (FSB) to assess the viability of different models for aggregating over-the-counter (OTC) derivatives data reported to trade repositories (TRs). The study evaluates three primary aggregation models—Option 1 (physically centralised), Option 2 (logically centralised), and Option 3 (individual authorities collecting raw data)—with the goal of providing a comprehensive analysis to support the FSB's decision-making on implementing a global aggregation mechanism.

The study emphasizes the importance of improving transparency, mitigating systemic risk, and protecting against market abuse in OTC derivatives markets, as agreed by G20 Leaders in 2009. It acknowledges that while Option 3 is currently in use, it is limited in its ability to meet the full range of authorities' data needs. Options 1 and 2 are considered more suitable for providing a comprehensive global view of the market, though the study does not definitively favor one over the other.

Main Objectives and Scope

• Objective: To evaluate the feasibility of different aggregation models for OTC derivatives data to support the FSB's mandate.
• Scope: The study covers the legal, data, and technological considerations of each model, and outlines the requirements for aggregated data to be useful to authorities.
• Approach: The study is high-level, focusing on the general effectiveness and scalability of each model rather than specific technological choices.

Aggregation Models

Option 1: Physically Centralised Model

• A central database collects and stores transaction and position data from TRs.
• The central facility provides aggregated data to authorities, performing quality checks, removing duplications, and anonymising data as needed.
• Data is stored centrally and accessible to authorities based on their access rights.

Option 2: Logically Centralised Model

• Data remains in local TR databases, with a central logical index used to locate and retrieve data.
• The central mechanism facilitates data retrieval on an "as needed" basis, without physically collecting or storing data.
• Similar to Option 1 in terms of data processing and anonymisation, but with data stored locally.

Option 3: Individual Authority Collection

• No central database or index; authorities collect raw data directly from TRs.
• Authorities perform data verification and processing individually.
• Limited ability to provide anonymised or aggregated data across multiple TRs.

Key Considerations

Legal Issues

• Legal obstacles to accessing TR data include differences in national laws and regulations.
• Legal challenges include data privacy, access rights, and the need for harmonisation of data standards.

Data and Technology

• The need for standardisation of data elements, such as the Legal Entity Identifier (LEI), Unique Transaction Identifier (UTI), and Unique Product Identifier (UPI), is critical for effective data aggregation.
• Data quality, management, and storage are essential to ensure accurate and reliable aggregated data.

Cost and Implementation

• Options 1 and 2 are expected to have similar costs, though distribution varies.
• Option 3 is likely to be more expensive globally due to the need for individual authorities to handle data processing and storage.

Recommendations

• Accelerate the development and adoption of global identifiers (LEI, UTI, UPI) to support data aggregation.
• Develop international guidance on data harmonisation to ensure consistency across TRs.
• Proceed with the development of a global aggregation mechanism in parallel with the introduction of identifiers.
• Conduct detailed legal, regulatory, and technical assessments before launching a formal project.

Conclusion

The study concludes that while Option 3 is currently in use, it is insufficient to meet the comprehensive data needs of authorities. Options 1 and 2 offer more robust solutions, though both require significant legal and technological coordination. A phased approach to implementing these models is recommended, starting with less comprehensive versions that can be expanded over time. The FSB is advised to take action to improve data standards and legal frameworks to enable the development of a globally effective aggregation mechanism.