2 May 2026 — Philosophy
The record should survive
Feeds get retired and their specifications quietly disappear with them. Public documentation of public markets should outlive the systems it describes.
Open Markets Initiative · 4 min read
Exchanges retire feeds. That is normal and healthy — protocols improve, hardware changes, a venue consolidates three market data products into one. What is less healthy is what happens to the documentation.
The specification comes down. The sample data comes down. The migration guide that explained how the old sequence numbering worked comes down with them. Within a year or two the only remaining description of a protocol that carried billions of messages a day is whatever happens to sit in private repositories at the firms that connected to it.
Why it matters after the fact
Retired protocols are not only of historical interest.
Packet captures outlive the feeds that produced them. Regulators, researchers, and firms reconstructing their own history all end up holding data they can no longer decode. A capture is just bytes without the specification that gives it structure, and "we have the data but lost the schema" is a genuinely common position to be in.
There is also a fairness argument. A venue's published specification was public when it mattered commercially. Withdrawing it later does not make the market safer; it just makes the record asymmetric, available to firms that archived it and unavailable to everyone else.
What OMI actually commits to
Two things, and deliberately not more.
Transparency. Public protocol specifications and test data should remain public and discoverable, including after the feed is gone. The Directory keeps them indexed and versioned rather than treating a retirement as a deletion.
Best practices. Everything in the ecosystem — loaders, compiler, generators — is built with automated builds, regression testing, and clean architecture. Documentation nobody can verify is not much better than no documentation.
Market-neutral, specifically
OMI does not take sides between exchanges, brokers, trading firms, or technology providers. It is not an advocacy project and has no position on market structure, fee models, or who should be allowed to connect to what.
The neutrality is not diplomatic hedging. It is what makes the archive usable: a record that anyone suspects of having a thesis is a record that gets argued with rather than cited. The goal is narrower and duller than advocacy — that the protocols which carried public markets remain legible to the public, and that the tooling for reading them is free.