After $5M of liquidations in an hour, does the move keep going?

282 historical matches·677 markets scanned·2026-05-15 to 2026-07-26

At a glance

What the historical record shows

Close to a coin flip after 30 minutes: 149 rose and 133 fell over 282 completed windows. No reference rate is available, so no lift is claimed.

Use this to decide what to investigate next. These historical frequencies do not establish a profitable trade.

▸ Replay one Edit this question →Check newer evidence ↓

Share text

After $5M of liquidations in an hour, does the move keep going?

282 matches · 677 markets · 2026-05-15 to 2026-07-26 UTC

Close to a coin flip after 30 minutes: 149 rose and 133 fell over 282 completed windows. No reference rate is available, so no lift is claimed.

Historical evidence, not a trading return or prediction.

Read the evidence and edit the question: https://edgedepth.com/research/reports/featureliqnotionalusd1h-5000000-34fc2ca4?utm_source=report-share&utm_medium=peer-share

Which markets does this cover?

All markets scanned by this study. No market or sector filter.

Results combine 677 scanned markets. They do not establish the outcome for an individual market or sector.

Asset-class and sector breakdowns are not included in this saved result. See research coverage →

Explore market coverage ↓

What is this report for?

Test a market observation against recorded history, compare what followed with the reference rate, then inspect examples before pursuing the idea.

Understand the setup and method ↓

Armed as a standing alert (definition #4), linked 2026-07-27: the live alert watches for exactly this definition; it and this page compile to the same ID.

Results by time periodCompare 30 minutes, 1 hour and longer periods, including larger moves

Choose a time period. “Closed” measures the price at its end; high and low points measure prices reached along the way. No same-scope reference is pinned to this report, so no ratio is claimed.

30 minutesView outcomes and counts

Outcome	After this setup	Reference minutes	How often vs reference
Closed 2% or more higher	1.8%5 of 282	Not available	Not available
Closed 2% or more lower	1.8%5 of 282	Not available	Not available
Closed 5% or more higher	0.0%0 of 282	Not available	Not available
Closed 5% or more lower	0.7%2 of 282	Not available	Not available
High point 5% or more above	not in ladder	Not available	Not available
Low point 5% or more below	not in ladder	Not available	Not available

1 hourView outcomes and counts

Outcome	After this setup	Reference minutes	How often vs reference
Closed 2% or more higher	not in ladder	Not available	Not available
Closed 2% or more lower	not in ladder	Not available	Not available
Closed 5% or more higher	not in ladder	Not available	Not available
Closed 5% or more lower	not in ladder	Not available	Not available
High point 5% or more above	0.0%0 of 282	Not available	Not available
Low point 5% or more below	1.1%3 of 282	Not available	Not available

4 hoursView outcomes and counts

Outcome	After this setup	Reference minutes	How often vs reference
Closed 2% or more higher	10.6%30 of 282	Not available	Not available
Closed 2% or more lower	6.0%17 of 282	Not available	Not available
Closed 5% or more higher	0.0%0 of 282	Not available	Not available
Closed 5% or more lower	0.4%1 of 282	Not available	Not available
High point 5% or more above	1.4%4 of 282	Not available	Not available
Low point 5% or more below	2.1%6 of 282	Not available	Not available

24 hoursView outcomes and counts

Outcome	After this setup	Reference minutes	How often vs reference
Closed 2% or more higher	16.3%46 of 282	Not available	Not available
Closed 2% or more lower	28.0%79 of 282	Not available	Not available
Closed 5% or more higher	2.8%8 of 282	Not available	Not available
Closed 5% or more lower	5.3%15 of 282	Not available	Not available
High point 5% or more above	9.2%26 of 282	Not available	Not available
Low point 5% or more below	16.7%47 of 282	Not available	Not available

72 hours: Not included in this saved report.

7 days: Not included in this saved report.

Average and median returns are not available in this report. These frequencies alone do not establish profitability.

When did it happen, and how spread out were returns?Frequency over time and the distribution of 30-minute returns

Occurrences per day · UTC · sum 282

Each bar counts matching setups on one UTC day. Look for clusters: many examples from a short period may reflect the same market episode.

Where price ended after 30 minutes

These three groups cover every complete outcome, including the moves smaller than 2%. Bar length shows the share of matches, not the size of the price move.

Fell 2% or more1.8%

5 of 282 matches

Between −2% and +2%96.5%

272 of 282 matches

Rose 2% or more1.8%

5 of 282 matches

0 incomplete outcomes excluded. This measures the price at 30 minutes, not the highest or lowest price reached on the way.

Detailed distribution: 14 return ranges

30-minute return distribution · 282 present · 0 absent. Each column counts outcomes in the labelled return range. The ranges have different widths; compare counts, not the shape as a probability density.

AUTHOR'S NOTE - OPINION, NOT MACHINE OUTPUT
"282 episodes of a $5,000,000-plus liquidation hour (deduped false-to-true anchors) over the full corpus: BTC 127, ETH 123, SOL 15 - a majors phenomenon. What follows is chop, not fireworks: 24h forward closed at or beyond +5% in 8 of 282 episodes and at or beyond -5% in 15 of 282; typical 24h MFE and MAE both sit in the 1-5% band. The cohort comparison (same document, cohort_result.v2) shows the tails are NARROWER than baseline, not wider: every other eligible bucket reached +5% in 24h 10.6% of the time and -5% 11.2% (of 41.9M buckets with a complete horizon). But the baseline spans 677 symbols while these episodes are majors-dominated, so much of that gap is composition, not calm. Cost line (flat assumption until measured spreads ship): ~2 bps half-spread on majors + 5 bps taker per side = ~14 bps round trip; deflate thresholds accordingly before reading a net edge. Verdict: "a big liquidation hour means fireworks ahead" does not survive the count. Descriptive only; not advice."

Results since publicationHave new matches appeared, and what happened after them?

Armed as a standing alert (definition #4, linked 2026-07-27): the live alert watches for exactly this definition. It and this page compile to the same ID.

Outcome	Original study, 2026-05-15 to 2026-07-26	Since publication, from 2026-07-27
Traded 5% or more above the match within 1h	0 of 282 (0.0%)	loading
Closed 2% or more higher at 24h	46 of 282 (16.3%)	loading
Closed 2% or more lower at 24h	79 of 282 (28.0%)	loading

Both periods are counted the same way. The later period is the same search on data recorded after publication, not a preregistered test. A missing outcome is never counted as zero.

Matches by marketMost matches: BTCUSDT 127 · ETHUSDT 123 · SOLUSDT 15677 markets itemized

Find where this setup appeared. These are occurrence counts, not performance rankings. Outcomes for individual markets are not included in this saved report.

Find a marketShowing 25 of 677 itemized markets

BTCUSDT127103,62112.3
ETHUSDT123103,62111.9
SOLUSDT15103,6211.4
SPCXUSDT573,3810.7
XRPUSDT4103,6210.4
XAGUSDT373,3810.4
DOGEUSDT2103,6210.2
AAVEUSDT173,3810.1
HYPEUSDT1103,6210.1
ZECUSDT1103,6210.1
0GUSDT073,3810.0
1000000BOBUSDT073,3810.0
1000000MOGUSDT073,3810.0
1000BONKUSDT073,3810.0
1000CATUSDT073,3810.0
1000CHEEMSUSDT073,3810.0
1000FLOKIUSDT073,3810.0
1000LUNCUSDT073,3810.0
1000PEPEUSDT0103,6210.0
1000RATSUSDT073,3810.0
1000SATSUSDT073,3810.0
1000SHIBUSDT073,3810.0
1000XECUSDT073,3810.0
1INCHUSDT073,3810.0
1MBABYDOGEUSDT073,3810.0

Examples and replaysFirst 50 of 282 matches

Examples, never the denominator; the counts above cover every match. Replay opens 30 minutes before the moment.

Sort the 50 examples shown here. This does not sort all study matches.

Replay

ETHUSDT2026-07-24 13:52−0.08%+0.34%−0.51%replay →