> ## Documentation Index
> Fetch the complete documentation index at: https://docs.dune.com/llms.txt
> Use this file to discover all available pages before exploring further.
# polymarket_polygon.open_interest_hourly
> Polymarket hourly open interest and traded volume per market, from the first mint to resolution.
export const TableSample = ({tableName, tableSchema}) => <>
;
The `polymarket_polygon.open_interest_hourly` table provides hourly open interest and traded volume per Polymarket market. Grain: one row per `(hour, market_id)`, with a row for every hour from the hour the market's outcome tokens were first minted to the hour it resolves. Open interest is a level at the end of the hour; volume is the flow within it.
Open interest is the outcome token supply, rebuilt from the CTF ERC-1155 mint and burn ledger on Polygon. History starts in April 2021. It is counted **per side**. On a standalone market the Yes and No sides are equal. On a member of a mutually exclusive (neg-risk) event, they diverge once No tokens are converted: converting No tokens of one outcome mints Yes tokens of every sibling outcome without adding collateral, so the No side drops and the siblings' Yes sides grow. `event_oi_contracts` gives the collateral locked behind the whole event.
Every row is the state at the end of its hour. A decided market has no open interest, so the resolution hour reads zero and the series stops there, even though tokens are still held: losing tokens are never redeemed, and winners redeem over weeks. Redemptions are in [`polymarket_polygon.market_actions`](/data-catalog/curated/prediction-markets/polymarket/market_actions). A market that opens and resolves within one hour shows volume but no open interest.
Combos are not included. A combo's collateral hangs on its legs, so it doesn't add to single-market open interest.
Don't sum `market_oi_yes_contracts` across the members of an event. Because conversions mint sibling Yes tokens without collateral, the sum can overstate the money at stake several times over. Don't sum `event_oi_contracts` across members either, because it is repeated on every member row. For total open interest, take `event_oi_contracts` once per event and hour, grouping by `COALESCE(event_id, market_id)` so that each standalone market counts as its own event. The example below does this.
Volume is kept here rather than on the candles because summing the Yes and No candles double-counts every Yes-versus-No match. Per-token prices are in [`polymarket_polygon.ohlcv_hourly`](/data-catalog/curated/prediction-markets/polymarket/ohlcv_hourly).
## Table Schema
| Column | Type | Description |
| - | - | - |
| `block_month` | `DATE` | First day of the UTC month of `hour`. Partition key |
| `hour` | `TIMESTAMP` | UTC hour the figures describe. Open interest is the level at the end of it, volume the flow within it |
| `market_id` | `VARCHAR` | Market condition ID. Joins `polymarket_polygon.market_details` on `condition_id` and `polymarket_polygon.ohlcv_hourly` on `market_id` |
| `event_id` | `VARCHAR` | Event this market is one outcome of, when the venue groups markets into a mutually exclusive set. NULL for a standalone market |
| `market_oi_yes_contracts` | `DOUBLE` | Yes tokens outstanding at the end of the hour. Zero in the hour the market resolves |
| `market_oi_no_contracts` | `DOUBLE` | No tokens outstanding at the end of the hour. Equal to the Yes side on a standalone market. Lower on an event member whose No tokens have been converted into Yes tokens of sibling outcomes. Zero in the hour the market resolves |
| `event_oi_contracts` | `DOUBLE` | Open contracts across the undecided members of the market's event at the end of the hour: the Yes tokens of one member plus the No tokens of every other member, which is the payout whichever outcome wins. A member that resolves early drops out of the sum from its resolution hour. Equal to the Yes side on a standalone market. One contract pays 1 unit of the quote stablecoin |
| `volume_contracts` | `DOUBLE` | Outcome tokens traded in the hour, both sides together, each match counted once |
| `volume_usd` | `DOUBLE` | Cash traded in the hour in the quote stablecoin, each match counted once from the aggressor's side |
| `trade_count` | `BIGINT` | Matches in the hour, both sides together |
| `_updated_at` | `TIMESTAMP` | When this row was last written by the pipeline |
## Table sample
## Query performance
`block_month` is the partition key. Always include a `block_month` or `hour` filter.
## Example query
```sql theme={null}
-- Total open interest at the end of each UTC day, last 30 days.
-- event_oi_contracts repeats on every market of an event, so take it once per event.
-- A market without an event_id is its own event.
WITH per_event AS (
SELECT
CAST(hour AS DATE) AS day,
COALESCE(event_id, market_id) AS event_key,
MAX(event_oi_contracts) AS open_interest_contracts
FROM polymarket_polygon.open_interest_hourly
WHERE block_month >= DATE_TRUNC('month', CURRENT_DATE - INTERVAL '30' DAY)
AND hour >= CURRENT_DATE - INTERVAL '30' DAY
AND HOUR(hour) = 23 -- the 23:00 row is the level at the end of the day
GROUP BY 1, 2
)
SELECT
day,
SUM(open_interest_contracts) AS open_interest_contracts
FROM per_event
GROUP BY 1
ORDER BY 1
```
This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.