All trades

What is this dataset?

Tick-level data is the most granular level of trading data, and contains every single trade that occurs or centralized and decentralized exchanges. The data is normalized and timestamped and contains information such as the price and volume of each trade. For DEX’s specifically, we also provide additional information on the user address, the blockchain, the pool address and transaction hash related to the trade.

Update frequencies

Channel	Update frequency
AWS	New CSV once a day
Azure	New CSV once a day
Google Cloud Platform	New CSV once a day
Snowflake	Data refreshes once a day
BigQuery	Data refreshes once a day

File structure details

• File Name: [exchange_name]_[data_feed_code]_trades_[date].csv

  • Files are created on a daily basis.

  • example: Binance V2_BTCUSD_PERP_trades_2023_02_13.csv

• Cut-off time: 00:00:00 UTC

  • We use the timestamp when exchanges send the order book data points through websocket or REST API in order to cut-off the data points between days.

• Column Delimeter: , (comma)

Column	Description	Example
id	Unique trade ID (unique to the exchange). In case the exchange does not provide an ID, we generate it ourselves.	1661990400274591751
exchange	Kaiko legacy slug for an exchange	bw
symbol	Data feed code in lower case	btcusd_perp
date	The timestamp provided by the exchange in Unix Timestamp (in milliseconds)	1676246400344
price	Executed price	21779.5
amount	Quantity of asset bought or sold (can be in base_asset, quote_asset or the number of contracts).	402
sell	Taker sell side	true
exchange BigQuery and Snowflake only	Exchange code. See Exchange codes	cbse
instrument BigQuery and Snowflake only	Exchange code. See Exchange codes	btc-usd