Borrows - Token Terminal Docs

In progress. This table is not served yet. The page documents it as it lands.

facts.lending_market_borrows combines the borrow events of every lending protocol we decode into one table: one row per borrow from a lending market. A borrow takes tokens out of the market against collateral posted elsewhere in the pool. lending_market_id keys each row to its market in the Registry.

Columns

Column	Type	Description
`block_timestamp`	`TIMESTAMP`	Time of the block containing the borrow. Partition column.
`block_number`	`INT64`	Number of the block containing the borrow.
`transaction_hash`	`STRING`	Hash of the transaction that emitted the borrow event.
`log_index`	`INT64`	Position of the borrow event in the block.
`unique_id`	`INT64`	Row key: fingerprint of `chain_id`, `block_number`, `transaction_hash`, `log_index`.
`chain_id`	`STRING`	Chain the borrow happened on.
`lending_market_id`	`STRING`	Market the borrow left; joins `dimensions.lending_markets`.
`project_id`	`STRING`	Protocol operating the market, such as `aave`.
`token_id`	`STRING`	Token borrowed: `{token_address}-{chain_id}`.
`token_address`	`STRING`	Contract address of the token borrowed.
`borrower_address`	`STRING`	Account the protocol’s event names as the borrower.
`on_behalf_of_address`	`STRING`	Account that carries the debt, when borrowed for another account.
`amount_raw`	`STRING`	Raw amount as text; a `uint256` exceeds every BigQuery number type.

Sample queries

The table is partitioned on block_timestamp. Bound that column in every query; without a bound the query reads the whole table.

One market, one day
Daily USD volume

Read one market’s borrows on a single day. Cast amount_raw to BIGNUMERIC and divide by 10^decimals from dimensions.tokens to get the token amount.

select
    borrows.block_timestamp,
    borrows.transaction_hash,
    borrows.borrower_address,
    cast(borrows.amount_raw as bignumeric) / pow(10, tokens.decimals) as amount
from `facts.lending_market_borrows` as borrows
join `dimensions.lending_markets` as markets
    using (lending_market_id)
join `dimensions.tokens` as tokens
    using (token_id)
where markets.project_id = 'morpho'
  and markets.underlying_symbol = 'WETH'
  and markets.chain_id = 'ethereum'
  and borrows.block_timestamp >= timestamp('2026-08-01')
  and borrows.block_timestamp < timestamp('2026-08-02')
limit 100

Sum daily borrow volume in USD per protocol. The table carries token amounts; USD comes from Prices.

select
    timestamp_trunc(borrows.block_timestamp, day) as day,
    borrows.project_id,
    round(sum(cast(borrows.amount_raw as bignumeric) / pow(10, tokens.decimals) * prices.price)) as borrowed_usd
from `facts.lending_market_borrows` as borrows
join `dimensions.tokens` as tokens
    using (token_id)
join `metrics.tokens_prices_daily` as prices
    on prices.token_id = borrows.token_id
    and prices.timestamp = timestamp_trunc(borrows.block_timestamp, day)
where borrows.block_timestamp >= timestamp('2026-08-01')
  and borrows.block_timestamp < timestamp('2026-08-08')
group by day, borrows.project_id
order by day, borrowed_usd desc

⌘I