Oracles

Paragon delivers accurate, timely index prices to Hyperliquid through a robust multi-source oracle infrastructure.

Price Sources

Primary: Pyth Network

Paragon sources prices primarily through a partnership with Pyth Network, a decentralized oracle providing high-frequency price feeds across 100+ blockchains. Pyth aggregates data from institutional-grade market data providers, delivering sub-second updates with broad asset coverage.

→ Pyth Documentation

Fallback: Hyperliquid-Inspired Mechanism

When Pyth data is unavailable, Paragon falls back to a Hyperliquid-inspired mechanism that polls spot mid-prices from CEX and DEX exchanges using a weighted median approach:

Exchange	Weight
Binance	3
OKX	2
Bybit	2
Kraken	1
Kucoin	1
Gate IO	1
MEXC	1
Hyperliquid	1

→ Hyperliquid Oracle Documentation

Mark Price

Paragon follows Hyperliquid's robust price indices model for mark prices. The mark price combines multiple data sources into a median calculation to mitigate manipulation risk:

1. Current Oracle Price

2. Current Oracle Price adjusted by a 150s EMA of the deviation between Hyperliquid's midpoint and the oracle

3. Internal exchange data — median of best bid, best ask, and last trade (supplied by HyperLiquid)

Mark price determines margin requirements, liquidation triggers, TP/SL execution, and unrealized PnL.

→ Hyperliquid Robust Price Indices

Data Pipeline

Pyth (primary) / Exchange Feeds (fallback) → Validation → Aggregation → Index Calculation → Hyperliquid

1. Collection: Price and supply data from Pyth (primary) or exchange feeds (fallback)

2. Validation: Outlier detection, staleness checks, sanity bounds

3. Aggregation: Weighted median across valid sources

4. Calculation: Index value computed per methodology

5. Delivery: Price updates posted to Hyperliquid via HIP-3 oracle interface

Index Methodology

True Float (Coming Soon)

Circulating supply data from industry vendors is often gatekept, opaque, and inconsistent—inaccuracies that compound across every downstream dependency built on this data. Paragon is building an open-source "True Float" methodology: a transparent, verifiable approach to circulating supply that anyone can audit and replicate.

Currently, circulating supply is aggregated from five data vendors: CoinMarketCap, CoinGecko, CoinPaprika, CryptoRank, and CoinStats.

Market Cap Indices (TOTAL2, OTHERS)

Index = Σ (Price_i × CirculatingSupply_i)

For each constituent asset:

• Price is the aggregated spot price in USD

• Circulating supply per True Float methodology above

Dominance Indices (BTC.D)

BTC.D = BTC_MarketCap / TOTAL

Where TOTAL = aggregate market cap of top 125 assets. Expressed as a decimal with four decimal precision.

Operational Parameters

Parameter	Value	Description
Price Updates	Sub-second	During active markets
Rebalancing	Daily	Constituent weight updates
Methodology Reviews	Quarterly	Or as needed
Uptime Target	99.99%	High availability commitment
Infrastructure	Multi-region	Redundant geographic distribution
Failover	Automatic	Backup data source fallback
Monitoring	24/7	Alerting and incident response

Transparency & Independent Verification

Market participants should be able to independently replicate and verify oracle prices. Paragon publishes all inputs required to reconstruct index values from scratch.

Published Data

Data Point	Description
Constituents	Full list of assets included in each index
Prices	Price per constituent with timestamp
Price Sources	Which data vendor provided each price (Pyth, exchange feeds)
Supply Figures	Circulating supply per constituent
Supply Vendors	Which vendor provided each supply figure
Timestamps	Exact time of each data point
Index Values	Computed index value with calculation timestamp

Provenance

Every data point includes its source attribution, allowing full traceability from raw inputs to final index value. This means anyone can:

1. Fetch the same constituent list

2. Pull prices from the stated sources at the stated timestamps

3. Pull supply figures from the stated vendors

4. Apply the published formula

5. Arrive at the same index value

Why This Matters

Opaque index calculations create trust dependencies. By publishing complete input data with provenance, Paragon enables trustless verification—you don't have to take our word for it.