Evidence before interpretation
Start with the source observation. State the reasoning separately. Do not present a hypothesis as a measured fact.
Independent Hyperliquid research built around clarity, provenance and context.
Markets move quickly. Understanding takes a closer look.
Liquid Data is a personal, independent research space for connecting market observations with the mechanics beneath them. It follows prices and positioning, the economics of HYPE, and the layers of the Hyperliquid ecosystem.
The aim is straightforward: make the evidence legible, explain the context, and leave the uncertainty visible.
Start with the source observation. State the reasoning separately. Do not present a hypothesis as a measured fact.
Name the provider, endpoint or document, unit, and collection time. Preserve source timestamps when they exist.
A current page can contain older observations. Retrieval time and source time are labeled separately when they differ.
Unavailable values remain a dash or an explicit unavailable state. We never infer a zero from an empty response.
A sample is labeled as a sample. Partial provider history is not called complete. Archive boundaries and gaps belong beside the result.
Liquid Data is an independent personal project, not an official Hyperliquid product. Research is read-only; this site does not hold funds, connect a signing wallet, or place orders.
Identify the source, field and unit.
Check dates, definitions and coverage.
Separate the evidence from its interpretation.
Public sources are a beginning, not a guarantee of completeness.
We check units, time boundaries, and field definitions before comparing observations. The market workspace reads public provider data. Protocol research reads a persistent daily archive with verified backfills, original timestamps, and visible gaps. Supply, burn and open-interest history begins where verified observations exist. Foundation notes carry a source review date and are not live protocol specifications.
Prices, funding, candles and wallet observations come from Hyperliquid’s public APIs. These feeds do not pass through hl.eco, Hyperdash or Hyperscreener.
Liquid Data saves the latest complete market catalog and native market snapshot on its server. If a refresh fails, it can return that saved observation for up to 24 hours, with its original retrieval time and an explicit stale-data warning.
Hyperliquid, Hypurrscan, DefiLlama and issuer websites remain upstream dependencies. Saved market snapshots preserve context during an outage; they cannot provide new prices. Other datasets retain their own coverage and unavailable states.
Live snapshots and the daily research archive have separate update paths. The archive records source observations through an authenticated collector, with the latest persisted run and coverage shown in Protocol research. A saved observation is not an executable quote or a guarantee of continuing source availability.
Liquid Data calculates native HYPE valuation from Hyperliquid inputs and is building its own recorded history. Other datasets still use the sources listed below; an independent index is not yet available for every metric.
| Dataset | Source | During an outage | Boundary |
|---|---|---|---|
| Native HYPE calculations | Hyperliquid spot metadata, order book and token details | Recorded values retain their original book time and supply receipt; stale current values are withheld | Calculated here: HYPE/USDC midpoint, spread and circulating value. Separate daily USDC history starts at observation; no USD peg or unobserved history is inferred. |
| Candle & capital archives | Hyperliquid + DefiLlama | Saved closed candles recover charts with original timestamps | Bounded capture: BTC/ETH/SOL/HYPE at 15m/1h/4h/1d, stablecoin history and up to five leading protocols per run. Missing periods remain gaps; protocol selection can change. Inspect saved coverage and records. |
| HYPE USD reference | CoinGecko; DefiLlama alternate delivery | Retained observations keep original provider and receipt times | Retained for comparisons with USD revenue and equities. Native supply valuation uses HYPE/USDC instead. Quotes older than five minutes are withheld; alternate delivery is not independent confirmation. |
| Protocol research archive | Hyperliquid, DefiLlama, Hypurrscan, issuer records, SEC and Nasdaq | Persisted observations and explicit gaps | Daily fees/revenue and holders have verified backfills. Supply, burns and OI start at observed coverage; ETF flows are estimates. |
| Trader tape | Hyperliquid wallet states; leaderboard and Hypurrscan discovery | Saved sample up to one hour, labeled with original times | Bounded public wallet sample. Open TWAP states and liquidation thresholds do not prove execution or future liquidations. |
| Order-book liquidity | Hyperliquid l2Book + WebSocket | 15-second snapshot fallback; original book timestamp | Bounded visible levels. Slippage is a static sweep excluding fees and order changes; no historical CEX comparison. |
| News & updates | Hyperliquid Announcements, Hyperliquid Policy Center, CoinDesk and HyperliquidNews | Per-source saved articles up to 7 days; dated checked selection remains separately labeled | Five-minute refresh while open; partial source coverage. Publication and retrieval dates are retained. A source announcement is not independent corroboration. |
| Markets & heatmap | Hyperliquid Info API + WebSocket | Last complete catalog, up to 24h | Native, HIP-3 and spot snapshots. No full historical market-wide archive. |
| Wallet research | Hyperliquid account, portfolio, fills, orders and TWAP slices | Previously loaded page data, labeled on failure | Account-specific, provider-bounded history. TWAP view covers up to 2,000 recent slices, not active-order completion or all historical executions. |
| HYPE supply & burns | Hyperliquid token details | Saved research observation, with original timestamp | Supply-state accounting plus recorded daily history. Locked versus vested-but-unclaimed reserves are not independently verified. |
| Holder statistics | Hypurrscan HyperCore index | Last complete snapshot, up to 24h | Reported holder counts and a ranked balance sample; not an independent address index. |
| Yield & lending | DefiLlama + Hyperliquid reserves | Last complete overview snapshot, up to 24h | Reported rates are not guaranteed returns. Pool history is fetched separately. |
| Auctions, bridges & stablecoins | Hypurrscan + DefiLlama | Last complete overview snapshot, up to 24h | Bounded bridge activity and provider-returned auction / stablecoin history. |
| Wall Street | Issuer disclosures & SEC documents | Dated verified snapshot | Selected treasuries and funds, not comprehensive institutional ownership. Reporting dates differ from market quote dates. |
| HIP-4 outcomes | Hyperliquid | Last complete overview snapshot, up to 24h | Outcome metadata; not full trading or settlement history. |
| Revenue, EVM & protocols | DefiLlama + HyperEVM RPC | Panel-specific prior observations | EVM fee estimates cover a small block sample, not lifetime network totals. No independent full-chain warehouse. |
Saved fallback retains original timestamps
Server fallback is available only after a successful complete snapshot has been stored. Expired or invalid snapshots are rejected. Stale data keeps its original retrieval time. No fallback can manufacture new observations during an upstream outage.
Hyperliquid-first analytics · migration in progress
HYPE midpoint, spread and circulating value now use Hyperliquid only. Independent fee, holder and event histories need verified indexing before their current providers can be retired. Issuer disclosures and external USD comparisons remain separately sourced. No additional quote provider is being activated.
Chart rendering & unavailable datasets
Charts use a pixel-accurate Three.js / WebGL layer when supported, with compatible canvas rendering otherwise. Exports use a fixed-resolution renderer so the evidence remains readable in an image. Complete liquidation history, market-wide live TWAP tracking and an independently operated historical index are not currently available.
Waiting for streaming market data.
Perpetual contexts and available spot reference prices stream. Individual figures retain their own receipt times. Protocol cards refresh public sources independently; historical charts and issuer reports keep their reporting dates.