Attributed by construction
Every figure carries the provider that produced it. Provenance is a property of the data model, not a footnote added at the end.
MarketCatalyst is a market-intelligence platform for people who have to form a view, not just retrieve a number. We assemble institutional-grade data across equities, estimates, filings and macro, hold it to a single standard of provenance, and deliver the interpretation alongside the evidence rather than leaving it as an exercise for the reader.
Scheduled jobs feeding the platform each session.
Market, fundamental, regulatory and macro.
Reported history behind every earnings view.
Rolling history under a published policy.
For most of the last decade the binding constraint in market research was access. Fundamentals were expensive, filings were awkward to parse, and real-time pricing was a subscription most participants could not justify. That constraint is gone. Data is abundant, cheap and, in aggregate, largely undifferentiated.
What did not scale with it is interpretation. A screen can tell you a position is down four percent, that consensus moved, that a filing landed after the close — and still leave the analyst to reconstruct, manually and from memory, which of those facts is the cause and which is the coincidence. That reconstruction is the actual work, and it remains almost entirely unautomated.
We treat that gap as the product. The platform is built so that the reasoning travels with the data: the same system that ingests the number is the one that explains it, against the same normalised record, with the source attached.
Pricing, fundamentals, estimates, ratings, ownership, corporate actions and macro series are drawn on their own cadences — intraday where the market demands it, on the filing clock where the regulator sets it.
Vendor payloads are reconciled onto a single company record with one identifier, one calendar and one set of units, so figures from different providers can be compared without silently disagreeing.
Surprise and revision history, relative strength, breadth, regime and clustered analyst conviction are computed in-house from the normalised record — never inherited from a vendor's own scoring.
A language model writes the read over that derived record. It explains evidence already present on the screen. It does not forecast, it does not rank, and it does not recommend.
Every figure carries the provider that produced it. Provenance is a property of the data model, not a footnote added at the end.
Ownership, insider activity and material events are read from EDGAR directly rather than from a summariser's paraphrase of it.
History is retained under explicit, auditable windows. Forward calendars are structurally excluded from expiry so scheduled events cannot be pruned.
Every endpoint and every upstream vendor is probed on an ongoing basis. Degradation is surfaced as a state of the platform, not discovered by a user.
MarketCatalyst is a research and information platform. It is deliberately not an advisory business, and the distinction is architectural rather than legal boilerplate.
Every read the platform writes is traceable to the figures underneath it — so you can check the reasoning, not just accept it.
MarketCatalyst is a data and research provider for informational and educational purposes only — not investment advice.