Funding vs account crowding
Contracts where the majority of accounts leans one way while the opposite side is quoted to pay funding.
Descriptive filters
Long accounts ≥ 55% with funding ≤ −0.005% per 8h, or long accounts ≤ 45% (short accounts ≥ 55%) with funding ≥ +0.005% per 8h. These thresholds select observations; they are not trading signals.
Common closed 5-minute target: . Each row shows its actual accounts time, recorded funding snapshot time and time difference.
Valid data, zero opposing directions: none of the 24 matched pairs meets these filters.
Long-account majority, funding negative
| Contract | Long / short accounts | Quoted payer | Funding / 8h | OI context | Accounts time UTC | Funding snapshot UTC | Skew |
|---|---|---|---|---|---|---|---|
| No matched pairs meet this filter. | |||||||
Short-account majority, funding positive
| Contract | Long / short accounts | Quoted payer | Funding / 8h | OI context | Accounts time UTC | Funding snapshot UTC | Skew |
|---|---|---|---|---|---|---|---|
| No matched pairs meet this filter. | |||||||
Excluded comparisons
Latest recorded funding is null or invalid: 25 · No accounts reading in the 20-minute window: 1
| Contract | Reason |
|---|---|
| Binance · BTCUSDT | Latest recorded funding is null or invalid |
| Binance · ETHUSDT | Latest recorded funding is null or invalid |
| Binance · SOLUSDT | Latest recorded funding is null or invalid |
| Binance · ZECUSDT | Latest recorded funding is null or invalid |
| Binance · BNBUSDT | Latest recorded funding is null or invalid |
| Binance · XRPUSDT | Latest recorded funding is null or invalid |
| Binance · HYPEUSDT | Latest recorded funding is null or invalid |
| Binance · NEARUSDT | Latest recorded funding is null or invalid |
| Binance · DOGEUSDT | Latest recorded funding is null or invalid |
| Binance · WLFIUSDT | No accounts reading in the 20-minute window |
| Binance · UNIUSDT | Latest recorded funding is null or invalid |
| Binance · SUIUSDT | Latest recorded funding is null or invalid |
| Binance · ENAUSDT | Latest recorded funding is null or invalid |
| Binance · LINKUSDT | Latest recorded funding is null or invalid |
| Binance · 1000PEPEUSDT | Latest recorded funding is null or invalid |
| Binance · TRXUSDT | Latest recorded funding is null or invalid |
| Binance · WLDUSDT | Latest recorded funding is null or invalid |
| Binance · PUMPUSDT | Latest recorded funding is null or invalid |
| Binance · ADAUSDT | Latest recorded funding is null or invalid |
| Binance · LTCUSDT | Latest recorded funding is null or invalid |
| Binance · TAOUSDT | Latest recorded funding is null or invalid |
| Binance · AAVEUSDT | Latest recorded funding is null or invalid |
| Binance · FILUSDT | Latest recorded funding is null or invalid |
| Binance · ONDOUSDT | Latest recorded funding is null or invalid |
| Binance · TRUMPUSDT | Latest recorded funding is null or invalid |
| Binance · JCTUSDT | Latest recorded funding is null or invalid |
What this means. One long and one short exist for every open contract. Account counts are not notional: each account counts once regardless of position size. Funding reflects the venue’s pricing mechanism. This comparison neither identifies large traders nor measures liquidation likelihood; neither side is necessarily wrong, and it is not a contrarian entry recommendation.
Matching. We select up to 50 Binance/Bybit USDT linear perpetual contracts by fresh engine OI ≥ $1M when the dataset is loaded. OI is context, not a weight. The newest all-accounts reading at or before the common target must still be no more than 20 minutes old at rendering. The newest funding snapshot at or before that reading must be within 5 minutes. Missing, null or invalid latest values are excluded; no older valid value, other venue, top trader accounts, top positions or taker flow fills a gap. Native symbols such as 1000PEPEUSDT are matched exactly.
Recorded snapshots. Funding is already an 8h-equivalent decimal: 0.00005 becomes 0.005% per 8h, with no extra interval scaling. Negative quotes mean shorts pay longs; positive quotes mean longs pay shorts. These are quotes, not settled payments. Funding timestamps identify recorded engine snapshot buckets, not necessarily the exchange’s original publication time. A recent bucket does not prove metadata freshness or synchronization within a bucket.
Within each group, rows sort by distance of the long-account share from 50%, then absolute funding quote, then venue/symbol. At most 50 rows appear across both groups. Source data comes from exchange public APIs through the existing engine and stored snapshots. The common dataset is cached for 120 seconds (30 seconds on failure); account ages are checked again on every render.