<< Back to articles list

Investing in OTC products is inherently complex. OTCStreaming's sophisticated pipeline (Figure 1) illustrates the multiple stages of data transformations. Initially, users require a ready-to-use solution, but they soon seek access to all parameters to assess the pipeline's sensitivity and potentially implement their own view using bespoke values. The OTCStreaming team has worked diligently to enhance the user experience and provide simple, comprehensive access to the entire system. Their objective was to enable non-tech-experts to utilize APIs, adjust pipeline parameters, and scale the pipeline to handle multiple data sources and products, beyond just the OTCStreaming-specific source, which involves the flow of credit data from dealers' emails. OTCStreaming introduces four key features allowing market practitioners to manage a complex pipeline without being tech experts:

• Query: Simplifies market data ingestion by providing a flexible structure for mapping user data with OTCStreaming symbology using advanced search algorithms, including new AI capabilities.
• Reference: Offers extensive coverage of OTC market securities and sophisticated price derivation algorithms for illiquid instruments.
• Correlation: Synchronizes market data with security descriptions and model parameters to create unified data objects for processing.
• Completion: Automates the identification and addition of missing market and security information, ensuring full transparency of the enrichment process.

These features empower users to efficiently handle complex data transformations and customize the pipeline to their needs.

Query

Market practitioners often deal with multiple security descriptions and are familiar with various symbologies, using any of them interchangeably. Integrating market data is challenging due to the diverse security identifiers across different sources. The Query feature provides a flexible structure that accompanies market prices, allowing securities to be specified using source-specific security descriptions. OTCStreaming processes these queries at the initial stage of the pipeline by matching them against the product referential using advanced search algorithms and AI capabilities. The system can create new securities when necessary and stores query resolutions for future reference. This approach simplifies market data ingestion by eliminating the need for multiple source mappings and builds a knowledge base that enhances future query processing.

Reference

OTCStreaming provides extensive coverage of OTC market securities across different liquidity levels. For illiquid instruments, the system uses sophisticated price derivation algorithms that leverage relationships with more liquid securities. In the CDS market, this includes mapping between off-the-run and on-the-run coupons, as well as connecting non-standard tenors to standard 5-year benchmarks. This systematic linking of market data enables users to implement custom price adjustments for illiquid securities as soon as the benchmarks move, resulting in proper SNAP (Synchronized Notification of Asynchronous Prices) market data generation on liquid and illiquid securities.

Correlation

OTCStreaming addresses the complexities of OTC markets through advanced correlation mechanisms. When market data arrives, such as an email containing both bond and CDS quotes, the system synchronizes this information with security descriptions and model parameters. The correlation process creates unified data objects that flow through the processing pipeline, where each stage is designed as a transformation function operating on these correlated objects.

Completion

Financial libraries often face challenges with mapping pricing functions to securities and gathering required market data. For example, derivative pricing needs zero-coupon curves, while bond basis calculations require credit default swap curves. OTCStreaming's pipeline engine automates this process. When users submit data, the system automatically identifies and adds missing market and security information to the correlation. Users can inspect these supplementary data points through the correlation, maintaining full transparency of the enrichment process.

These abstractions simplify the design of the OTCStreaming pipeline, making it easier to understand and providing users with various options to customize the default data processing mechanism. To manage these abstractions, the OTCStreaming team has adopted a highly adaptable record description using JSON records, encompassing financial objects and market data to complex correlations, covering both processing and storage. This blend of abstraction and flexible data description allows OTCStreaming to introduce new features and products much faster than competitors.

<< Back to articles list