Experimental data research

Whale edge: signal performance study

A falsifiable study of returns after whale-positioning imbalances, scored only against snapshots that were actually archived later.

Experimental section. Every page publishes its method, coverage and limitations; history accumulates only from real timestamped snapshots and is never back-filled.

Response generated: 2026-07-12 20:26 UTCHistory accumulating: 1 archived observation · 2026-07-12 to 2026-07-12

Updates: daily archive snapshot after the nightly data pipeline (~03:30 UTC); on weekdays an extra market reading at 15:30 UTC. Live tables refresh continuously.

Qualifying biases today

Markets that currently pass the study threshold: 60%+ of tracked whale notional on one side. Each becomes a ledger observation scored against the +1 and +7 day snapshots.

Long bias

  • STRC75% · 18/11 · $3.45M
  • ENS66% · 2/16 · $1.31M
  • MNT66% · 10/15 · $1.29M
  • PENDLE66% · 28/24 · $3.21M
  • CFX65% · 3/16 · $2.23M
  • LINEA65% · 9/17 · $1.01M
  • JPY62% · 7/10 · $6.11M
  • NBIS62% · 36/32 · $30.45M

Short bias

  • SPX90% · 9/34 · $1.2M
  • NFLX90% · 7/10 · $1.36M
  • PENGU88% · 15/31 · $1.01M
  • PURRDAT81% · 14/13 · $3.49M
  • TIA81% · 17/22 · $1.06M
  • ARB77% · 22/35 · $3.31M
  • ONDO75% · 16/36 · $3.13M
  • MINIMAX74% · 6/11 · $2.39M

Display filter matches the whales page: at least 15 wallets and $1M in open positions. The ledger itself records every qualifying bias. · Full whale overweight page →

Research snapshot

Key measurements

Evaluated horizons
0

Observed records

50 records shown.

“Return” and “Hit” fill in one and seven days after each bias observation, once the archive has the follow-up snapshots — early rows simply have not matured yet.

First evaluations as the archive matures: 1 day — 2026-07-13 · 7 days — 2026-07-19.

Date Ticker Whale bias Long share Price Return 1d Hit 1d Return 7d Hit 7d
2026-07-12GBPBUY
100.0%
$1.342
2026-07-12JP225BUY
100.0%
$68,661
2026-07-12BANANASELL
0.0%
$3.467
2026-07-12BIRDSELL
0.0%
$3.116
2026-07-12GMESELL
1.4%
$21.793
2026-07-12IBMSELL
6.5%
$286.82
2026-07-12SNXBUY
91.7%
$0.2422
2026-07-12REZSELL
8.9%
$0.00259
2026-07-12VINEBUY
91.1%
$0.00925
2026-07-12MANTASELL
8.9%
$0.05894
2026-07-12DKNGSELL
9.4%
$26.335
2026-07-12COSTSELL
9.5%
$915.46
2026-07-12SPXSELL
9.8%
$0.37623
2026-07-12ZENSELL
10.1%
$4.239
2026-07-12NFLXSELL
10.4%
$73.262
2026-07-12PALLADIUMSELL
11.0%
$1,267
2026-07-12PENGUSELL
11.9%
$0.00607
2026-07-12BSVSELL
12.1%
$13.469
2026-07-12DYMSELL
13.4%
$0.01562
2026-07-12RIVNSELL
13.6%
$17.215
2026-07-12ASMLSELL
13.9%
$1,793.9
2026-07-12KASSELL
16.4%
$0.02871
2026-07-12TURBOSELL
17.5%
$0.00084
2026-07-12BERASELL
17.8%
$0.19234
2026-07-12SUPERSELL
17.9%
$0.08658
2026-07-12PURRDATSELL
19.0%
$7.808
2026-07-12TIASELL
19.1%
$0.40199
2026-07-12GMTSELL
19.4%
$0.00766
2026-07-12PLATINUMSELL
19.5%
$1,623
2026-07-12NOWSELL
19.6%
$106.69
2026-07-12QNTSELL
19.9%
$70.139
2026-07-12WDCSELL
20.0%
$577.72
2026-07-12LAYERSELL
21.6%
$0.0661
2026-07-12EBAYBUY
78.2%
$117.68
2026-07-12ZMSELL
21.8%
$89.846
2026-07-12DASHSELL
21.8%
$34.885
2026-07-12NXPCSELL
22.5%
$0.26117
2026-07-12ARBSELL
22.7%
$0.09381
2026-07-12EWZSELL
23.7%
$36.781
2026-07-12ONDOSELL
25.0%
$0.32518
2026-07-12STRCBUY
74.7%
$88.313
2026-07-12BABYSELL
25.3%
$0.01304
2026-07-12GOATBUY
74.5%
$0.01383
2026-07-12MINIMAXSELL
25.5%
$35.088
2026-07-12EWTSELL
25.9%
$106.67
2026-07-12KR200SELL
27.0%
$1,192
2026-07-12TRBBUY
73.0%
$15.066
2026-07-12TSLASELL
27.3%
$407.85
2026-07-12FOGOSELL
28.1%
$0.0097
2026-07-12ICPSELL
28.5%
$2.256

Sources and coverage

Primary market data: Hyperliquid public API. Historical comparisons use only archived Spottrench snapshots that actually exist.

Calculation method

Each qualifying imbalance is timestamped before its forward-return window. Results use the saved due-date price; missing future snapshots remain unscored rather than backfilled.

Snapshot method

Freezes a long bias at 60%+ whale notional share and a short bias at 40%-; evaluates it only on exact +1 and +7 day snapshots.

Frequently asked questions

What exactly does this study test?

Whether a strong whale imbalance — 60%+ of tracked notional on one side of a market — predicts the direction of price moves one and seven days later. The bias is frozen the day it is observed, before the outcome is known.

How is a “hit” counted?

A hit means the sign of the return at exactly +1 or +7 days matched the bias direction. If the archive is missing the follow-up snapshot, the observation stays unscored — nothing is interpolated.

When will results appear?

The first 1-day scores arrive a day after the first archived observation, 7-day scores a week after. The summary hit rate is withheld until at least five scored observations exist.

If the hit rate is high, should I copy whales?

Not by itself. The sample can be selection-biased, trading costs and funding are not modeled, and correlation in a sample is not causation. Treat this as a public experiment, not a strategy.

Known limitations

  • The tracked-wallet sample may be selection-biased. Costs, slippage, funding and overlapping observations can materially reduce any apparent edge.
  • Neutral observations are not forced into long or short labels.
  • Missing exact future snapshots remain null and are excluded from hit-rate calculations.
  • Association in this tracked-wallet sample is not evidence that whales caused later returns.
  • At least five evaluated horizon observations are required before status becomes ok.

Experimental statistics are not a recommendation or financial advice. Correlation does not establish causation, tracked wallets are only a sample, and past relationships may disappear.