Wallet-flow regime signals (monthly, market-wide)
Progold_wallet_flow_1mo monthly, one market-wide row per month (no coin dimension)
gold_wallet_flow_1mo is a monthly, market-wide Hyperliquid regime dataset: one row per month aggregating every wallet across every coin (converted to USD) into whale vs retail net taker flow, new-wallet inflow, profit breadth and volume concentration. Whale net flow turns negative as tops distribute and positive on genuine rallies; new-wallet inflow tends to lead. Pro only, via the Tessera API.
A market-wide (coinless) monthly aggregate of Hyperliquid wallet behaviour — the regime layer on top of the per-wallet metrics. Every wallet's volume and net taker flow are converted to USD and summed across all coins, then wallets are split into a top-1%-by-volume whale cohort and a bottom-90% retail cohort (contemporaneous, in-month). The row captures whale vs retail net flow, the normalised retail-minus-whale discriminator, the share of new wallets, profit breadth, maker toxicity, and volume concentration. Pro only.
What you'd use it for: A regime overlay: gate or size a strategy by whether whales are accumulating or distributing, confirm turns with new-wallet inflow, or blend a market-structure feature into your own model.
Descriptive market structure, not a coin-level signal — one row per month, no coin dimension. In the studied window whale net flow was strongly negative through the top and decline and turned positive only on the bear-market rally, while new-wallet inflow led regime turns. Cohorts are defined on the month being measured (contemporaneous), so read it as a concurrent/near-lead regime read, not a look-ahead forecast.
Data dictionary
Month
| Column | Type | What it means |
|---|---|---|
month_start | timestamp[us] | The month this market snapshot covers (its first day, UTC). How it's computed: First day of the month (UTC). |
Participation
| Column | Type | What it means |
|---|---|---|
active_wallets | uint32 | How many distinct wallets traded anywhere on the venue this month — overall participation. How it's computed: Distinct wallets that traded in the month. |
new_wallet_frac nullable | float64 | Share of this month's active wallets that are appearing for the very first time. Rising new-wallet inflow tends to lead regime turns (it ran hot before the rally, cold at the bear trough). Null for the first month of available history — there is no prior month to measure "new" against. How it's computed: Fraction of active wallets whose lifetime first month is this month. Null for the earliest month of history (no prior month to measure new wallets against). |
Flow (USD)
| Column | Type | What it means |
|---|---|---|
total_volume_usd | float64 | Total USD notional traded across all wallets and coins — the denominator the flow cohorts are normalised against. How it's computed: Total taker+maker USD volume across all wallets and coins. |
whale_net_flow_usd | float64 | Net taker flow (USD, positive = net buying) of the top 1% of wallets by monthly USD volume. Negative as smart money distributes into strength; positive when it steps in to buy — the sharpest regime tell. How it's computed: Net taker flow (USD, +buy/-sell) of the top-1% wallets by USD volume. |
retail_net_flow_usd | float64 | Net taker flow (USD, positive = net buying) of the bottom 90% of wallets by volume — the retail cohort. How it's computed: Net taker flow (USD, +buy/-sell) of the bottom-90% wallets by USD volume. |
retail_minus_whale_norm | float64 | (retail net flow − whale net flow) ÷ total volume: high when retail is buying while whales sell (distribution), low/negative when whales buy while retail capitulates (accumulation). The core smart-vs-dumb discriminator. How it's computed: (retail_net_flow_usd - whale_net_flow_usd) / total_volume_usd — the regime discriminator. |
Quality & concentration
| Column | Type | What it means |
|---|---|---|
frac_profitable | float64 | Fraction of active wallets that booked positive realized PnL this month — breadth of profitability (a coincident regime confirmer). How it's computed: Fraction of active wallets with positive realized PnL in the month. |
maker_markout_wmean nullable | float64 | Volume-weighted average 30-minute maker markout across wallets — how adversely the market moves against resting liquidity (venue toxicity). How it's computed: USD-maker-volume-weighted mean 30m maker markout across wallets (null if no maker volume). |
whale_volume_share | float64 | Share of total USD volume done by the top-1% whale cohort — how concentrated activity is in the largest wallets. How it's computed: Top-1% cohort's share of total USD volume (concentration). |
coins_covered | uint32 | How many coins had a monthly price available for the USD conversion — a coverage/quality indicator for the aggregate. How it's computed: Number of coins with a monthly price used in the USD conversion. |