??? ??? ??? ??? ??? ??? ?? ??? ??? ??? ??? (???, SOL-PERP ??? BTC-PERP) ??? ??? ??? ??? ??? ???. ??? ??? ??? ??? ??? ???, ??? ??? ?? ??? ??? ??? ???, ??? ??? ??? ??? (Binance, Bybit) ?? ???-??? ??? ??? (Hyperliquid) ??? ??? ??? ??? ???.

??? ???, ??? ??? ??? ??? ???, ??? ??? ??? ??? (OI) ?? ??? ??? ??? ???, ??? ??? ??? ??? ??? ???. ?? ??? ??? ??? ??? ??? ?? ??? ??? ??? ??? ??? ??? ???.

??? ??? ???: 1-??? ??? 8-??? ???

??? ??? ??? ??? ??? ??? ??? ??? ?? ??? ??? ??? ?? ??? ??? (clamping).

historical funding rate alignment: 1-hour vs 8-hour timestamp mappingChronological UTC bar grid showing how Hyperliquid settles funding every hour (00:00–08:00 UTC) while 8-hour CEX contracts settle only at discrete 00:00 and 08:00 UTC snapshots. DISCRETE FUNDING SNAPSHOT ALIGNMENT ACROSS UTC TIMESTAMPS 1-Hour Clock (HL) 8 discrete events 01:0002:00 03:0004:00 05:0006:00 07:0008:00 8-Hour Clock (CEX) 1 snapshot at 08:00 00:01 – 07:59 UTC: $0.00 Funding Settled on 8h Venue 08:00 Common Backtest Bug: Applying an unadjusted 8-hour rate to every 1-hour bar overstates carry by 8.0x Only credit or debit funding when the simulated position is open across the exact settlement timestamp Verify contract interval adjustments · some CEX contracts dynamically switch from 8h to 4h or 1h clocks
Chronological UTC bar grid showing how Hyperliquid settles funding every hour (00:00–08:00 UTC) while 8-hour CEX contracts settle only at discrete 00:00 and 08:00 UTC snapshots. ?????????????????? ?????????????????? ?? ???? ???????????????? ?????????????? ?????????????????? ??????????????

??? ??? (Binance, Bybit) ??? ??? ??? ??? 8 ??? (?? 00:00, 08:00 ?? 16:00 UTC). ?? ??? ??? ???, Hyperliquid ??? ?? ??? ??? ??? (24 ??? ??? ?? ???).

??? ??? ??? ??? ??? ??? ??? ??? ???
Hyperliquid 1 ??? (???) 1-??? TWAP ??? 1 ??? ??4.00% ?? ???
Binance USD???-M 8 ??? (???) 8-??? TWAP ??? ??? ??0.75% ??? ??2.00% ??? 8??
Bybit USDT Perp 8 ??? (???) 8-??? TWAP ??? ??? ??0.75% ??? ??1.50% ??? 8??

??? ??? ??? ??? ??? ??? ??? ??? ??? Hyperliquid ??? ??? ???, ??? ??? ??? ??? ??? ???. ?? ??? ??? ??? ??? 8-??? ??? ??? Binance ??? ??? ???, ??? ??? ??? ??? ???.

??? ??? ??? ??? ???, ??? ??? ??? ??? ??? ??? ??? ???. ??? ??? ??? ??? ??? ??? ?? ??? ??? ??? ??? ?? ???-???.

??? ??? ???-??? ?? ??? ???-???

??? ??? ??? ??? ??? ??? ??? ??? ??? ??? ?? ??? ??? ???, ??? ??? ??? ??? ??? ?? ??? ???.

negative funding rate and spot-perp basis discount dynamicsDiagram showing how a perpetual mark price trading at a persistent discount below the spot oracle index drives negative funding payments from shorts to longs. PERPETUAL MARK TRADING AT DISCOUNT TO SPOT INDEX (P_perp < P_index) Spot Oracle Index Price = $64,000 Negative Basis Discount = −$200 (−0.3125% below Spot) Perpetual Mark Price = $63,800 · Settled Funding Rate = −0.050% per 8h SHORT POSITION HOLDERS Pay Negative Funding Debit −$50 per $100k notional every 8h Cash Transfer LONG POSITION HOLDERS Receive Funding Rebate Credit +$50 per $100k notional every 8h Negative funding incentivizes arbitrageurs to buy perp and pushes short sellers toward cover decisions
Diagram showing how a perpetual mark price trading at a persistent discount below the spot oracle index drives negative funding payments from shorts to longs. ?????????????????? ?????????????????? ?? ???? ???????????????? ?????????????? ?????????????????? ??????????????

??? ??? ??? ??? ??? ??? ?? ???, ??? ??? ??? ??? ??? ??? ??? ??? ???. ??? ??? ??? ??? ???:

  1. ??? ??? ???: ??? ??? ??? ??? ??? ??? ??? ??? ??? ?? ??? ???.
  2. ??? ??? ???: ???-??? ??? ??? ??? ??? ??? ?? ??? ??? ??? ??? ??? ??? ??? ?? ??? ??? ???, ??? ??? ???.
  3. ??? ?? ???-???: ??? ??? ??? ??? ??? ??? ???, ??? ??? ?? ??? ??? ??? ??? ??? ??? ??? ???. ??? ??? ??? ?? ??? ??? ???-??? ??? ??? ???.

??? ??? ??? ??? ??? ??? ???

??? ??? ??? ??? ??? ??? ??? ???, ??? ??? ??? ??? ??? ??? ??? ??? ??? ???.

open interest confirmation filter in backtest signal logicCausal point-in-time signal diagram showing how lagging open interest to closed bar t-1 prevents lookahead leakage while filtering breakouts by OI expansion. POINT-IN-TIME OPEN INTEREST FILTERING (NO LOOKAHEAD LEAKAGE) CLOSED BAR (t − 1) • Bar Close Price confirmed • Snapshot OI_(t-1) finalized • Bar CVD_(t-1) finalized Signal_t = f(Close, OI_t-1) BRANCH A: PRICE BREAKOUT + OI EXPANDING (> +5%) ✓ New capital entering trend → Model Next-Bar Open Entry (Bar t) BRANCH B: PRICE BREAKOUT + OI CONTRACTING (< −3%) ✗ Short-covering exhaustion only → Filter Out / Stand Aside Lookahead Audit: Never use Bar t closing OI to trigger an entry at Bar t Open Confirm timestamp conventions (bar-open vs bar-close tags) in your historical OI data provider before testing
Causal point-in-time signal diagram showing how lagging open interest to closed bar t-1 prevents lookahead leakage while filtering breakouts by OI expansion. ?????????????????? ?????????????????? ?? ???? ???????????????? ?????????????? ?????????????????? ??????????????

??? ??? ??? ??? ??? ??? ??? ??? ??? ??? ???:

  • ??? ??? (??? ??? OI): ??? ??? ??? ?? ??? ??? ??? ??? Binance, Bybit ?? Hyperliquid. ??? ??? ??? ??? ?? ??? ??? ??? ???.
  • ??? / ???-??? (??? OI): ??? ??? ???, ??? ??? ??? ??? ??? ???. ??? ??? ??? ??? ??? ?? ??? ???-???, ?? ??? ??? ???.
  • ??? ??? (Hyperliquid ??? CeFi): ??? ??? ??? Hyperliquid ??? ??? ??? ??? OI ??? ??? ???. ??? ??? ??? ??? ??? ??? ??? ??? ??? ??? ???.

??? ??? ??? ??? ??? ??? ?? ??? ???-??? ??? ??? ?? ??? ??? ??? ??? ?? ??? ??? ?? ???.

??? ??? ??? ??? ???

???, ??? ??? ??? ??? ???, ??? ??? ??? ??? ??? ???, ??? ??? ?? ??? ??? ???. Binance ?? Bybit ??? ??? ??? ??? ??? ??? (???, BNB), ??? ??? Hyperliquid ??? ??? ??? ?? ??? USDC.

??? ??? ??? ??? ??? ??? ??? ??? ?? PnL. ??? ??? ??? ??? ?? ??? ??? ??? ?? ??? ??? ??? SuperQuant.