Margin Town compresses market signals for speed while documenting formulas, fallbacks, liquidation direction and estimate limits.
Domestic KRW ÷ (global USDT × USDT/USD × USD/KRW) - 1. USDT keeps its own premium instead of becoming the conversion baseline. Non-USDT assets require current active SPOT quotes from at least two independent overseas exchanges that agree within the consensus guard; the median of the agreeing prices becomes the reference. Single-source and disputed symbols are withheld from the scanner.
Public Binance, Bybit, Bitget and OKX sources are separated from fallback and temporarily limited states wherever the source quality can be identified.
Only disclosed listing information is used. No countdown is invented when launchTime/listTime is unavailable.
The label means the position that was force-closed. Rally → more SHORT liquidation pressure; selloff → more LONG liquidation pressure.
Bitget history notional is a price × amount comparison estimate.
Korean pages use KST; English pages use UTC for Yesterday/Today/Tomorrow classification. Calendar sourcing and coverage can change with provider conditions.
This is a fast market context layer, not the exchange ledger. Confirm critical prices, fees and status at the official venue before execution.
Open Market Scanner →Freshness before completeness. A smaller verified dataset is preferred to a larger table containing stale or disputed references. Consensus before extremes. Cross-exchange identity and price agreement are checked before an extreme premium is treated as display-eligible. Visible failure states. Upstream restrictions, fallback use and estimated values should remain distinguishable from normal primary-source data.
If you find a reproducible data or methodology issue, send the page URL, symbol, displayed value and approximate time to admin@margin.town. Reports that affect the shared engine are reviewed across both language versions before the next release.
