Skip to content

Order Book Data

The FSI accelerator contains an Order Book Pipeline which is used to generate Order Book data from Realtime Level 2 Data.

How the Order Book Pipeline Works:

  1. The *-orderbook-* SP subscribes to OrderbookUpdates table from the *fsi-data* RT South.

    • OrderbookUpdates:
      • The OrderbookUpdates schema contains all information Level 2 Quote that is required to build Order Book data including:
        • side = BID | ASK
        • action = Add | Update | Delete

  2. Using the OrderbookUpdates data the Orderbook SP populates two tables:

    • OrderbookSnapshots:
      • The OrderbookSnapshots schema is a table of market-by-price orderbook data which represents the depth of an instrument using enlisted bids and asks bucketed by the configured snapshot size.
    • Depth:
      • The Depth schema is a table of market-by-price orderbook data which represents the depth of an order using values in an explicit depth column.

Both the OrderbookSnapshots and Depth tables are published back to the *fsi-data* RT South to be stored by the FSI SM.

Top of Book (TOB) Logic

To help customers keep accurate order books on their own systems, some feeds use both top-of-book (TOB) indicators as a way to flush invalid orderIDs from the orderbook.

A new TOB order is represented by containing a non-null value in the tobMarker column in the L2Quote schema.

If a new TOB is received then any orders that are higher than the TOB are removed from the book, i.e. - For the bid-side of the book orders with price greater than the new TOB price are removed. - For the ask-side of the book orders with price less than the new TOB price are removed.