Early access is open. Free sample datasets are live now. Download a sample

Data product · Early access

Dark Pool and Off-Exchange Trade Data

Every US stock trade that printed away from a lit exchange, with its size, the reporting facility, the NBBO at that moment and flags for blocks and late reports, so you can see where large size changed hands.

Coverage

What you get

Item Detail
Product code STK-DARKPOOL
Coverage US stocks and ETFs: trades reported through FINRA trade reporting facilities (TRFs)
Granularity One row per print
History Same depth as our stock trades history
Delivery REST and WebSocket; in files through our stock trades datasets
Update speed As prints are reported

A large share of US stock volume trades off-exchange: in dark pools, through wholesalers and in negotiated blocks. Those trades still get reported, just to a FINRA TRF instead of an exchange tape. Retail traders watch the big prints for hidden institutional interest. We filter them out of our full trade feed and tag what matters, so you aren't paying a separate vendor for a subset of data you'd already have.

Schema

Schema

Field Type Required Description
symbol string yes Our ticker
timestamp string yes Report time, ISO 8601 UTC
executed_timestamp string optional Execution time if different (late prints)
price number yes Trade price
size number yes Shares
notional_value number yes price times size
trf_id string yes Reporting facility
bid_price number optional NBBO bid at report
ask_price number optional NBBO ask at report
conditions array of strings optional Sale conditions
is_late boolean yes Reported 15 or more minutes after execution
is_block boolean optional Size at or above our block threshold

Example record (illustrative values):

{
  "symbol": "AMD",
  "timestamp": "2026-11-18T19:47:12.553Z",
  "executed_timestamp": "2026-11-18T19:47:12.553Z",
  "price": 141.5,
  "size": 250000,
  "notional_value": 35375000,
  "trf_id": "FINRA_NYSE_TRF",
  "bid_price": 141.48,
  "ask_price": 141.52,
  "conditions": ["@"],
  "is_late": false,
  "is_block": true
}

Sample data

Sample data

There's no free dark pool sample yet, because US trade data needs a licence. A sample file is available on request: give us a ticker and a date and we'll send that day's off-exchange prints in this schema.

Use cases

Use cases

Block print levels. Filter is_block true for a ticker over the last month and group by price. Prices where several large blocks traded are the "dark pool levels" many traders mark on their charts.

Where in the spread did it print? Compare price with bid_price and ask_price. A big print at or above the ask leans buyer-initiated, one at the bid leans seller-initiated, and one at the midpoint tells you little about direction.

Off-exchange share. Sum size for the day and divide by total volume from our bars. A jump in the off-exchange share on a quiet price day can mean size is moving without showing on lit books.

Cleaning late prints. Drop or re-time rows where is_late is true, using executed_timestamp. Late prints arrive at old prices and can look like sudden spikes on an intraday chart.

Code

Code examples

Example: blocks in AMD for one day.

curl "https://api.marketdata.co/v1/stocks/darkpool?symbol=AMD&date=2026-11-18&is_block=true" \
  -H "Authorization: Bearer $MARKETDATA_API_KEY"

Example: Python, notional by price level over a month.

import os
import requests
import pandas as pd

resp = requests.get(
    "https://api.marketdata.co/v1/stocks/darkpool",
    params={"symbol": "AMD", "start": "2026-10-19", "end": "2026-11-18", "is_block": "true"},
    headers={"Authorization": f"Bearer {os.environ['MARKETDATA_API_KEY']}"},
)
resp.raise_for_status()
prints = pd.DataFrame(resp.json()["results"])
prints["level"] = prints["price"].round(0)
levels = prints.groupby("level")["notional_value"].sum().sort_values(ascending=False)
print(levels.head(10))

Pricing

Pricing

Plan Dark pool prints
Pro, $79/mo Off-exchange prints from our trade feed, with block and late flags
Business, $249/mo Commercial internal use

As files, off-exchange prints come inside our stock tick data datasets ($15 per symbol-year); filter on the off-exchange flag. Price-level clustering is planned as a later add-on. Among the providers we studied, dark pool data costs $99 to $149.99 a month, median $99.

Side by side

How we compare

Only 3 of the 50 providers we studied sell dark pool data as a product, all at $99 a month or more, and one is web only with no API. Because off-exchange prints are already part of the full trade tape, we can include them in Pro at $79 and give you the same rows in the trade files you buy. Same symbol, same UTC timestamps, no second vendor.

FAQ

Frequently asked questions

Is every off-exchange trade a dark pool trade?

No. Trades reported to a FINRA TRF include dark pool (ATS) executions, wholesaler fills of retail orders and negotiated blocks. The TRF report doesn't always say which venue type it was, so we label prints as off-exchange and let is_block and size guide you.

What does "late" mean?

is_late is true when the trade was reported 15 or more minutes after it executed. executed_timestamp holds the real execution time.

What time zone are timestamps in?

UTC, in ISO 8601. The regular US session, 9:30 to 16:00 New York time, is 14:30 to 21:00 UTC in winter.

Can you tell me who bought?

No. TRF reports don't name the buyer or seller. Anyone who says a print shows "institutions buying" is reading the price and size, not a name.

Is it real-time?

Prints are planned to stream as reported on Pro, subject to our trade data licence. Some prints arrive late by rule, which is what is_late flags.

What formats are available?

JSON from the API, and CSV or Parquet through the stock trades datasets.

Keep reading