Data methodology
Every number in the terminal is an observation with a source, a calculation with a published formula, or a labelled estimate. This page is generated from the same definitions the pipeline runs.
Metrics
Loading
TidePerpEvery number in the terminal is an observation with a source, a calculation with a published formula, or a labelled estimate. This page is generated from the same definitions the pipeline runs.
Each instrument has a canonical code VENUE:PERP:BASE:SETTLEMENT:STYLE, for example HYPERCORE:PERP:BTC:USDC:LINEAR. Builder-deployed DEXs on HyperCore are separate venues (HYPERCORE-XYZ) because they have their own order books, fees and operators.
Instruments on the same underlying form a comparison group (BTC_PERPETUAL). Within a group, DIRECT instruments share the settlement asset and are ranked as is; NORMALIZED instruments settle in a different USD-pegged asset and are ranked with a disclosure; NON_COMPARABLE instruments differ in contract terms and are listed but never ranked. Aggregates (volume, open interest, market share) include DIRECT and NORMALIZED instruments on HyperCore-native and Elysium-native venues only.
Symbols such as kPEPE keep their source symbol; the unit multiplier (1,000) is recorded in the specification so prices and sizes are never silently rescaled.
| Metric | Formula | Units | Notes |
|---|---|---|---|
| Order-book imbalance | (D_bid − D_ask) ÷ (D_bid + D_ask) | ratio | note: Descriptive measure of visible depth, not evidence of directional intent. |
| Order-book depth within band | Σ price_i × size_i over bid levels with price ≥ mid × (1 − b) and ask levels with price ≤ mid × (1 + b) | USD | bands_bps: [10,50,100,200] source_depth: at most 20 levels per side (HyperCore l2Book) |
| Hypothetical execution cost | Walk visible levels best-first until filled; VWAP = Σ p_i q_i ÷ Σ q_i; slippage_bps = |VWAP − mid| ÷ mid × 10,000; total = half-spread cost + impact + fees | USD / bps | impact: |VWAP − best opposite| × filled quantity spread_cost: |best opposite − mid| × filled quantity reference_price: mid of the snapshot insufficient_depth: returns INSUFFICIENT_OBSERVED_DEPTH; never extrapolates beyond observed levels |
| Annualized funding rate |
| Detector | Description | Parameters | Enabled |
|---|---|---|---|
| Funding divergence | Hourly-normalized funding differs between comparable instruments or against relayed external venues. | normalize_to_seconds=3600min_divergence_hourly=0.0001 | |
| Funding extreme | Current funding rate beyond the 30-day percentile band. | baseline_days=30percentile_low=2.5percentile_high=97.5min_abs_rate_hourly=0.00005 | |
| Liquidity deterioration | Depth within the 100 bps band fell versus its 24h median. | band_bps=100max_ratio=0.4baseline_hours=24min_baseline_depth_usd=100000 |
HyperCore networks
Documented limits respected by the ingester
{
"docsVerifiedAt": "2026-11-07",
"docsUrl": "https://hyperliquid.gitbook.io/hyperliquid-docs/for-developers/api/rate-limits-and-user-limits",
"restWeightPerMinute": 1200,
"infoWeights": {
"l2Book": 2,
"allMids": 2,
"clearinghouseState": 2,
"orderStatus": 2,
"spotClearinghouseState": 2,
"exchangeStatus": 2,
"userRole": 60
},
"defaultInfoWeight": 20,
"perItemWeights": {
"recentTrades": 20,
"fundingHistory": 20,
"candleSnapshot": 60
},
"maxTimeRangeElements": 500,
"maxCandles": 5000,
"maxBookLevels": 20,
"maxWsConnections": 10,
"maxNewWsConnectionsPerMinute": 30,
"maxWsSubscriptions": 1000,
"maxWsMessagesPerMinute": 2000,
"maxWsInflightPosts": 100,
"wsIdleTimeoutSeconds": 60,
"hyperEvmPublicRpcRequestsPerMinute": 100
}| Venue | Tier | Maker | Taker | Effective | Verification | Source |
|---|---|---|---|---|---|---|
| HYPERCORE | BASE | 1.50 bps | 4.50 bps | 2026-11-07 | docs (opens hyperliquid.gitbook.io in a new tab) |
| (1 + r)^n − 1 with n = 31,536,000 ÷ funding_interval_seconds (365-day year) |
| fraction |
note: Native rate and interval are always stored alongside. |
| Cumulative funding | Π (1 + r_i) − 1 over the selected window | fraction |
| Funding divergence | normalize(r_a, interval_a → 1h) − normalize(r_b, interval_b → 1h) with normalize(r, f→t) = (1 + r)^(t/f) − 1 | fraction per hour | comparability: only DIRECT or NORMALIZED instruments of the same comparison group |
| Market share | V_venue ÷ Σ V_comparable_venues × 100 | % | universe: instruments of the same comparison group and execution domain with DIRECT or NORMALIZED comparability, trailing 24h |
| Potential net carry | expected_funding − entry_fees − exit_fees − slippage − financing | USD | sign: positive funding is paid by longs to shorts disclosure: estimate under stated assumptions; not a guarantee of profitability |
| Open interest change (24H) | OI_now − OI_24h_ago; percentage = difference ÷ OI_24h_ago × 100 | USD / % | reference: nearest market snapshot at or before now − 24h |
| Open interest (USD) | OI_base × mark_price (base-denominated linear contracts); OI_contracts × multiplier × mark_price otherwise | USD | note: Outstanding exposure, not turnover. Never doubled for both sides. |
| Price change (24H) | (mark_price − prev_day_price) ÷ prev_day_price × 100 | % | reference: source-reported previous-day price (prevDayPx) |
| Bid/ask spread | (ask − bid) ÷ mid × 10,000 with mid = (bid + ask) ÷ 2 | bps | note: Not calculated when either side is missing or the book is crossed/invalid. |
| 24H volume (USD) | source-reported rolling 24h notional per instrument (HyperCore dayNtlVlm); aggregates sum it over the comparable universe | USD | source: the venue’s own rolling 24h notional, read with every ticker; TidePerp’s trade capture (Σ price × quantity × multiplier) is reconciled against it hourly and reported separately as trade coverage window: rolling 24h |
Analytics library notes
VOLUME_NOTIONAL_USD v1: V_usd = Σ_i P_i × Q_i × M_i × C_i (price × quantity × contract multiplier × quote-to-USD factor)
OPEN_INTEREST_USD v1: BASE: OI_usd = OI_base × P_mark; CONTRACTS: OI_usd = OI_contracts × M × P_mark
MARKET_SHARE_PCT v1: Share_i = V_i / Σ_j V_j × 100 over the comparison universe
SPREAD_BPS v1: Mid = (Bid + Ask) / 2; Spread_bps = (Ask − Bid) / Mid × 10,000
DEPTH_BAND_USD v1: D_bid(b) = Σ P_i × Q_i for P_i ≥ Mid × (1 − b); D_ask(b) = Σ P_i × Q_i for P_i ≤ Mid × (1 + b); b = band_bps / 10,000
BOOK_IMBALANCE v1: Imbalance = (D_bid − D_ask) / (D_bid + D_ask)
EXECUTION_COST v1: Walk the opposite side best-first until filled; VWAP = Σ P_i Q_i / Σ Q_i; spread = |best_opposite − reference| × Q; impact = |VWAP − best_opposite| × Q; fee = notional × fee_bps / 10,000; total = spread + impact + fee; slippage_bps = |VWAP − reference| / reference × 10,000
FUNDING_ANNUALIZED v1: n = 31,536,000 / interval_seconds; COMPOUND: (1 + r)^n − 1; SIMPLE: r × n
FUNDING_CUMULATIVE v1: F_cumulative = Π_i (1 + r_i) − 1
FUNDING_NORMALIZED v1: r_to = (1 + r)^(to_seconds / from_seconds) − 1
FUNDING_DIVERGENCE v1: Divergence = normalize(r_a, interval_a → T) − normalize(r_b, interval_b → T), default T = 3600 s
NET_CARRY v1: NetCarry = ExpectedFunding − EntryFees − ExitFees − Slippage − Financing; ExpectedFunding = ±r × notional × periods (+ SHORT, − LONG); fees and slippage = notional × bps / 10,000; Financing = notional × annual_rate × horizon_s / 31,536,000
ROBUST_Z_SCORE v1: z = (x − median(baseline)) / (1.4826 × MAD(baseline)); MAD = median(|x_i − median|)
| Market stress | Several separately visible stress indicators active for the same market. | window_minutes=60min_active_signals=3 |
| OI contraction | Open interest falling sharply relative to 24h ago. | min_oi_usd=1000000lookback_hours=24min_change_pct=-15 |
| OI expansion | Open interest rising sharply relative to 24h ago and its 30-day percentile. | min_oi_usd=1000000lookback_hours=24min_change_pct=15percentile_threshold=95 |
| Price dislocation | Mark price deviates from oracle price beyond a threshold. | min_deviation_bps=50 |
| Spread expansion | Spread widened versus its 24h median. | min_ratio=3baseline_hours=24min_spread_bps=5 |
| Volume anomaly | Hourly notional volume versus the rolling 7-day median using robust (MAD) deviation. | min_buckets=48z_threshold=3.5baseline_hours=168bucket_seconds=3600min_notional_usd=250000 |
Elysium networks
Elysium facts (official documentation)
{
"docsVerifiedAt": "2026-11-07",
"blockTimeMs": 100,
"gasPerSecondTarget": 300000000,
"dataAvailability": "ANYTRUST_DAC_WITH_PARENT_FALLBACK",
"sequencer": "SINGLE_SEQUENCER_PRIORITY_FEE_ORDERING",
"validation": "PERMISSIONED_AT_LAUNCH",
"upgradeDependent": [
"HYPERCORE_MARKET_DATA_PRECOMPILE",
"ELYSIUM_CORE_WRITER",
"KEEPER_LANE"
],
"mainnetStatus": "PRE_LAUNCH_CHAIN_ID_AND_EXPLORER_PUBLISHED_AT_LAUNCH"
}