# Canonical: https://www.uddalak.xyz/hyperliquid-academy/article/110-hyperliquid-public-data-market-research
# Author: Uddalak Das
# Last source: https://www.uddalak.xyz/hyperliquid-academy/article/110-hyperliquid-public-data-market-research

---
title: "How to Use Hyperliquid's Public Data for Market Research"
description: "Hyperliquid's public books, trades, funding, open interest, account state, and protocol metadata can support reproducible market research if definitions, timestamps, and survivorship are handled carefully."
pillar: "Markets & Trading"
track: "Markets & Trading"
cluster: "Markets & Trading"
level: "Advanced"
primary_icp: ["data analyst", "trader"]
secondary_icp: ["institutional decision-maker"]
content_type: "guide"
reading_time: "3 min"
last_reviewed: "2026-09-07"
last_verified: "2026-09-07"
volatile: true
prerequisites: ["95-how-hyperliquid-market-data-is-structured"]
related: ["103-open-interest-volume-tvl-stablecoin-balances", "106-how-to-measure-slippage-and-market-impact"]
---

# How to Use Hyperliquid's Public Data for Market Research

Hyperliquid exposes unusually rich market state through public APIs and onchain infrastructure. That makes it possible to research liquidity, funding, positioning, trader behavior, vaults, and market structure without buying a proprietary dataset. The hard part is not downloading data. It is preserving enough context to avoid misleading conclusions.

## Define the question first

“Study Hyperliquid volume” is not a research question. Examples of testable questions are: Does positive funding predict short-horizon returns? How does depth change before liquidation clusters? Does maker spread widen nonlinearly with volatility? How long does HIP-3 basis persist?

The question determines which feeds and sampling frequency are required.

## Collect raw state

Persist market metadata, L2 books or selected depth bands, trades, candles, funding, open interest, mark/oracle prices, and relevant account/vault data. Store timestamps in one canonical timezone and retain the original source identifier.

If an endpoint has history limits, run a continuous collector rather than assuming the API will supply an unlimited retrospective dataset later.

## Avoid look-ahead

Backtests must use information that was actually available at the decision timestamp. A final candle close, revised metadata, or end-of-hour funding payment cannot be used to make a trade earlier in that interval.

Order-book research also needs realistic latency and queue assumptions.

## Handle changing markets

Symbols list, delist, migrate, and change parameters. A dataset containing only markets that survived to the end introduces survivorship bias. Save market-status and specification history where possible.

Protocol upgrades should be treated as regime boundaries in long time-series studies.

## Publish reproducibly

State the endpoint, field definitions, sample period, missing-data handling, filtering, and code assumptions. Distinguish observation from causal claim.

For Academy research pieces, link the raw query or code where practical so a reader can reproduce the result instead of trusting a chart image.

## How this connects

Start with [How Hyperliquid Market Data Is Structured](/hyperliquid-academy/article/95-how-hyperliquid-market-data-is-structured) if the underlying concept is unfamiliar. Related reading: [Open Interest vs Volume vs TVL vs Stablecoin Balances](/hyperliquid-academy/article/103-open-interest-volume-tvl-stablecoin-balances), [How to Measure Slippage and Market Impact on Hyperliquid](/hyperliquid-academy/article/106-how-to-measure-slippage-and-market-impact).

## Sources

- https://hyperliquid.gitbook.io/hyperliquid-docs/for-developers/api/info-endpoint
- https://hyperliquid.gitbook.io/hyperliquid-docs/for-developers/api/websocket/subscriptions
- https://hyperliquid.gitbook.io/hyperliquid-docs/for-developers/api/rate-limits-and-user-limits