Smart-Money Positioning Score
TRADER+A −100 … +100 view of the latest tracked Hyperliquid open-position notional within the selected window. Positive means the quality- and recency-weighted book leans long; negative means it leans short. It does not predict price.
| Symbol | Flow | Score | Direction | Wallets | Sample | Long $ | Short $ |
|---|---|---|---|---|---|---|---|
| Loading… | |||||||
For each wallet and symbol, the service keeps the most recent open-position snapshot inside the selected 1–168 hour window. Long notional is positive and short notional is negative.
The base wallet weight uses a 14-day snapshot-quality proxy: 70% from the share of open-position snapshot rows with positive unrealized P&L and 30% from a saturating positive-P&L term. Wallets with fewer than 8 rows are pulled toward the 0.25 floor; all weights are bounded to 0.25–1.00. For wallets covered by the offline realized-fill cache, the weight is blended 65% toward that realized quality; flagged wallets are floored.
Position contributions decay exponentially with age using the code’s 12-hour time constant — exp(−age / 12h), an e-folding time, not a half-life. The final score is 100 × weighted net notional ÷ weighted absolute notional.
- ±100 means the weighted book is almost entirely one-sided; it does not mean high expected return.
- Long $ and Short $ are raw open notional; Score uses quality and recency weights.
- Wallets is the number of unique contributors for that symbol.
- Sample labels are thin at ≤4 wallets, ok at 5–15, and rich above 15.
- Coverage is Hyperliquid only.
Source: the Hyperliquid whale_positions snapshot table plus an offline realized-fill quality cache for covered wallets. Results are cached for up to 2 minutes. The page refreshes on load or when the window changes; it does not poll automatically.