Historical Exchange IndexExchange IndexA quiet registry of crypto exchanges, active and gone.

Data quality

Quality and coverage

Inspect where the registry is strong, where evidence or archives are thin, and which reviewed records need further strengthening.

Entities1069
Events1097
Evidence4026
Generated2026-09-01
Entity archive coverage95%

entities with an archived URL

High-confidence entities58%

share of all reviewed entities

Average evidence per entity3.8

total evidence records divided by entities

High-reliability evidence51%

2072 of 4026 evidence records

Archived evidence share89%

3596 / 4026 evidence items

Entities with 0 evidence0

0% of entities

Entity confidence

Reviewed confidence classification on exchange entity records.

High620 · 58%
Medium438 · 41%
Low11 · 1%

Evidence reliability

Reviewed reliability classification on public evidence records.

High2072 · 51%
Medium1903 · 47%
Low51 · 1%

Evidence depth by entity

Number of evidence records associated with each reviewed exchange entity.

zero0 · 0%
1 source2 · 0%
2–4 sources773 · 72%
5+ sources294 · 28%

Record freshness

Time since the entity record was last reviewed, using last_verified_at.

0–30 days122 · 11%
31–90 days755 · 71%
91–180 days192 · 18%

Coverage

Archive, date, origin, domain, and confidence coverage with explicit denominators.

MetricValueDenominator / note
Archive coverage951013 / 1069 entities
Dead-side archive coverage100231 / 231 dead-side
Active-side archive coverage93782 / 838 active-side
High-confidence share58620 / 1069 entities
Launch date known44468 / 1069 entities
Death date known99229 / 231 dead-side
Origin known1001069 / 1069 entities
Original domain known951013 / 1069 entities

Evidence source types

Composition of the public evidence layer by source category.

Official statement1576 · 39%
Official blog236 · 6%
Official social43 · 1%
Archive capture179 · 4%
News article661 · 16%
Court document9 · 0%
Regulatory notice161 · 4%
Database reference1096 · 27%
Community reference60 · 1%
Other5 · 0%

Evidence claim scope

What kind of registry claim each evidence record is intended to support.

Entity1214 · 30%
Event865 · 21%
Status1184 · 29%
Death reason54 · 1%
Launch date398 · 10%
Death date68 · 2%
URL history198 · 5%
Ownership45 · 1%

Selected missing-field counts

Visible selected field gaps. A retained unknown value is not automatically an error.

MetricValueDenominator / note
Missing launch date60156% of entities
Dead-side missing death date21% of dead-side
Missing origin00% of entities
Missing original domain565% of entities

Completeness indicators

Selected public coverage indicators. These do not form a hidden quality score.

MetricValueDenominator / note
Aliases present78834 / 1069 entities
Notes present1001067 / 1069 entities
Summary present1001069 / 1069 entities
Archived URL present951013 / 1069 entities
Dead-side with 2+ evidence100230 / 231 dead-side
Entities with 0 evidence00% of entities
Entities with 1 evidence20% of entities

Metric definitions

Definitions and denominators used to interpret this public summary.

Entity confidence
The reviewed confidence classification stored on each exchange entity record.
All reviewed public entities.
Evidence reliability
The reviewed reliability classification stored on each evidence record.
All reviewed public evidence records.
Evidence depth
How many evidence records are associated with each exchange entity.
All reviewed public entities.
Archive coverage
Whether an entity record includes an archived URL reference. It does not measure the completeness of every historical page.
The entity population named in each metric note.
Record freshness
Elapsed time since the entity last_verified_at date, grouped into review-recency bands.
All reviewed public entities.
Unknown-field rates
Visible missing values in selected public entity fields. Unknown values are not automatically treated as errors.
All entities or dead-side entities, as stated in each metric note.
This summary is generated from reviewed public HEI records. Internal monitoring findings, private research notes, unresolved candidate queues, and operator-only repair priorities are not published here.