Retail and institutional data feeds compared

Consolidated tape data is a centrally assembled stream of trades and quotes aggregated across markets. Direct exchange feeds are venue-native products that can expose full depth of book or order-by-order updates. Which one a study needs follows from whether the question is about cross-market activity or about one venue's book.

For researchers working with U.S. equity market data, the practical difference between consolidated tape data and direct exchange feeds is not just distribution path. It is a structural difference in who assembles the data, what fields are included, how much order-book state is visible, and how precisely event timing can be represented.

The Consolidated Tape Association oversees the Securities Information Processor for New York Stock Exchange and NYSE American listed securities, and the consolidated tape is a real-time stream that combines last-sale and quotation information from the major U.S. securities markets. [1] The consolidated product is an aggregation layer across venues, not a venue-native feed: it is assembled centrally and carries trade and quote information across exchanges rather than the full native message stream of any single matching engine.

By contrast, NYSE's market data catalogue lists exchange-provided proprietary feeds such as Integrated Feed, OpenBook, Trades, BBO, Order Imbalances and ArcaBook. The descriptions matter: OpenBook provides a real-time view of the full depth of limit orders in the NYSE and NYSE American books, and ArcaBook provides order-by-order depth for NYSE Arca. [2] Direct feeds are supplied by each exchange from its own book, and can expose native depth-of-book information, including order-level updates, that sits outside the scope of any consolidated product.

That structure is exactly why depth differs. A consolidated tape aggregates top-level trade and quote information from multiple venues into one common stream, so it does not represent the full resting order inventory of each venue's limit order book. Exchange-native feeds are generated from the venue's own book, which is what makes full-depth and order-by-order products possible at all. For research questions about queue evolution, displayed liquidity, or the precise reconstruction of one venue's book, the two data classes are materially different things.

Timing precision follows the same logic, with one point worth stating carefully. NYSE's proprietary historical products are timestamped in nanoseconds. [2] Consolidated products carry their own timestamping conventions, which vary by product and generation, so a study that depends on fine-grained event ordering should verify the actual timestamp specification of the exact product in hand rather than assume a rule about feed classes. The safe general statement is narrower than folklore: venue-native historical products can carry nanosecond stamps, and the researcher must check what their consolidated source actually provides.

For research use, the consequence is straightforward. If the study needs cross-market last-sale and quote aggregation, consolidated tape data matches the question, because it summarizes trading and quoting activity across venues in a single stream. If the study needs to reconstruct venue-specific book dynamics, observe full displayed depth, or analyze order-by-order changes in a particular exchange book, exchange-native feeds such as OpenBook or ArcaBook are the relevant class, because the book content only exists there.

A small example makes the boundary concrete. Suppose a researcher wants to study how displayed liquidity changes immediately before and after a trade on one venue. With consolidated inputs, they can observe reported trades and quotes aggregated across markets, but not the sequence of depth changes inside the venue's book. With a direct feed carrying full depth or order-by-order messages, they can watch price levels and displayed size change around the event. The question decides the feed: cross-market activity points to the consolidated stream, and the inside of one venue's book points to that venue's own feed. [3]

Drafted with AI assistance from cited sources. Reviewed and approved by Sonar Sciences Quant & Research Team.