12 March 2026 — Origins
Origins of the Open Markets Initiative
Financial markets are the most sophisticated real-world distributed systems we have, and their rules of operation are unevenly documented. OMI exists to make those rules explicit, machine-readable, and equally available to everyone.
Open Markets Initiative · 3 min read
Financial markets are highly evolved distributed systems. Decades of competition and collaboration around price discovery have produced an explosion of protocols for moving information, coordinating participants, and resolving state across networks.
OMI is a computer science research initiative focused on distributed systems. We study financial markets because they contain a broad range of the most sophisticated real-world protocols.
As a starting point for analyzing these protocols, the rules of operation need to be clear and equally accessible. Today, exchanges vary in the quality of their implementation and documentation and in their policies for communicating changes and answering questions. That creates an information asymmetry favoring the largest firms, which have giant engineering teams and exclusive exchange relationships.
OMI reduces that asymmetry by creating a consistent, transparent layer covering the diversity of exchange connectivity and implementations. We believe machine-readable definitions can describe protocols more precisely than PDFs alone, making the rules of operation explicit, testable, and equally available to everyone.
We have translated a range of financial protocols into complete and open definitions in the form of Kaitai Structs, DFDL, XML, P4, Spicy, and QuickFIX dictionaries.
Along the way, we're exploring better ways to model protocols themselves, precisely enough that they can generate working code and full systems. We use our protocol definitions to generate Wireshark dissectors that are relied on across the trading industry. This helps validate their accuracy and usefulness. For the same purpose we are generating open-source protocol artifacts for C, C#, C++, Java, Python, Rust and more languages.
As our research progresses toward a platform capable of modeling the full diversity of financial protocols, we believe the same ideas can help tackle some of the hardest communication problems in distributed systems more broadly.
We welcome contributions from engineers in trading firms and exchanges, protocol experts, code generation enthusiasts, network engineers, and anyone who wants to work on the cutting edge of distributed systems.