Findings · 2026-11-12
Following whale trades: 1,694 events, no edge
Whale alerts promise a look at what the big money is doing. We rebuilt them from raw exchange trades on six major altcoins and followed every one.
The claim. “When whales buy, follow them.”
What we found. Before fees the result was zero. After fees it was a small, steady loss.
What we measured
We went to the source: every individual futures trade on SOL, XRP, DOGE, AVAX, LINK and ADA, sampled every fifth day from January 2022 to October 2026. A whale trade was any single trade of $250,000 or more.
An event was an hour in which whales bought (or sold) at least $1 million more than they sold (or bought), and that imbalance was at least 3% of everything traded in the hour.
What happened
Following the whales' direction for the next four hours returned -0.12% per trade after fees on the unseen part of the data (1,541 events). Before fees the result was about zero, so the whole loss is the cost of trading. The whales' direction carried no information about the next four hours.
Measuring each coin against Bitcoin over 24 hours, to remove the effect of the whole market moving, gave -0.04%. Also nothing.
We also tested a third-party whale alert feed for the week we had it: 426 alerts. Following them lost 1.44% against Bitcoin on average. One week is too short to be a verdict, but it points the same way.
Why it does not work
A large trade always has someone equally large on the other side. Seeing the trade tells you a big buyer and a big seller agreed on a price, and nothing about which of them was right.
How it was tested
- Binance futures trade-by-trade data, six coins, 2,118 coin-days, January 2022 to October 2026.
- Enter at the close of the event hour. Costs: 0.107% taker fees plus real funding.
- Rules locked 10 November 2026 before any result. Data split at 20 September 2023.
Research on past data, not financial advice. Every test we run, failures included, is on the Proving Ground.