Skip to content
Genesis Ark

07 / Market data

Available

One normalized view of every connected order book

Qubit's market data module connects to each venue's feeds, normalizes symbols, prices and sizes into one model, detects gaps and resynchronizes, and builds a consolidated order book across venues. Routing, risk, pricing and TCA all read from it, and your own systems can too.

Layer
Standalone, read-only
Works on its own
Yes
Interfaces
WebSocket, REST, FIX
Fig. 1. A consolidated order book with each level tagged by its source venue. Illustrative composition with sample data.

01 / The problem

Every exchange speaks its own dialect

One venue calls it BTCUSDT, another BTC-USDT, a third XBT/USDT. One sends full book snapshots, another sends incremental updates with sequence numbers, a third drops messages under load without telling you. Prices and sizes come with different precisions. Building a correct book for one venue is work; keeping correct books for many, through disconnects and venue changes, is a permanent engineering cost.

02 / What it does

What the module provides

  1. 01

    Normalized order books

    Level 2 books per venue, built from snapshots and incremental updates, in one format with one instrument model.

  2. 02

    Trades and top of book

    Public trades and best bid and offer per venue, timestamped by the venue and on receipt.

  3. 03

    Consolidated order book

    One book per instrument across selected venues, with each level showing which venue it came from.

  4. 04

    Composite reference price

    A single reference price per instrument built from selected venues, with rules for excluding stale or outlying venues. Used for risk collars, marks and pricing.

  5. 05

    Reference data

    Instruments, tick sizes, lot sizes, minimum notional, trading status and fee schedules per venue, kept up to date.

  6. 06

    Gap detection and resync

    Sequence gaps, stale feeds and crossed books are detected, and the affected book is rebuilt from a fresh snapshot.

  7. 07

    Capture and replay

    Recorded data can be replayed for TCA, testing and incident review.

03 / How it works

From venue feed to normalized book

Fig. 2. The stages below, in order. Each one writes to the audit trail.
  1. 01

    Connect

    Qubit subscribes to each venue's public feeds, by WebSocket where available.

  2. 02

    Normalize

    Symbols map to Qubit instruments; prices and sizes convert to a common precision; timestamps are kept as both venue time and receipt time.

  3. 03

    Validate

    Sequence numbers, checksums where the venue provides them and book integrity are checked on every update.

  4. 04

    Consolidate

    Per-venue books merge into a consolidated book and feed the composite reference price.

  5. 05

    Distribute

    Internal modules and your applications subscribe over WebSocket, or query snapshots over REST.

04 / Specifics

Technical specifics

Data types

L1 (best bid and offer), L2 (aggregated depth by price level), public trades, instrument reference data, venue status. L3 order-by-order data where a venue provides it.

Book depth

Configurable per instrument and venue, up to the depth the venue publishes.

Timestamps

Venue event time and Qubit receipt time on every message, so latency between venue and receipt can be monitored in your own environment.

Feed health

Last update time, message rates, gap counts and resync events per venue and instrument, exposed in the console and over the API.

Distribution

WebSocket subscriptions per instrument and data type; REST for snapshots and reference data; FIX market data (MarketDataRequest 35=V, MarketDataSnapshotFullRefresh 35=W, MarketDataIncrementalRefresh 35=X).

Licensing note

Venue data is used under each venue's own terms. Redistribution of venue data to third parties may need permission from the venue; that is between you and the venue.

05 / Controls and audit

What you configure, and what gets written down

Configurable

  • Venues per instrument in the consolidated book
  • Venues and exclusion rules for the composite reference price
  • Book depth per instrument and venue
  • Silence threshold before a feed is flagged stale
  • Subscriptions per instrument and data type
  • What is recorded, for how long and where

Logged to the audit trail

  • sequence gap detected · venue, instrument
  • feed flagged stale · last update time
  • book rebuilt from snapshot
  • crossed book detected
  • venue excluded from or restored to the composite price
  • reference data change · tick size, lot size, status

06 / Where it fits

Works on its own
Yes. It needs no exchange accounts for public data, and it cannot place orders.
Needs the core
No. Inside the platform, it feeds smart order routing, pre-trade price collars, client pricing, portfolio marks and TCA.
Who uses it
Quant and trading teams, brokers and OTC desks building prices, fintech apps, risk teams that need consistent marks.

FAQ

Questions we get asked

  • Is this a market data vendor feed?

    No. Qubit connects to venues directly from your infrastructure and normalizes their feeds for your own use. It does not resell data. If you already use a data vendor, Qubit can use it as an additional source.

  • How do you handle a venue whose feed goes stale?

    The venue is flagged stale after a configurable silence, its levels drop out of the consolidated book and the composite price, and Qubit resubscribes and rebuilds the book from a snapshot.

  • Can we use the consolidated book in our own applications?

    Yes. Subscribe over WebSocket or query over REST. The format is the same one Qubit's own modules use.

  • Do you store historical data?

    Data can be recorded in your environment for TCA, testing and replay. How much you keep, and where, is your choice and depends on your storage.

Stop maintaining exchange feeds yourself

A scoping call with an engineer, then a demo on sample data that mirrors your venues, desks and limits.