Research · Altcoins
ICP drawdowns and recoveries, measured against Bitcoin's
Last reviewed 2026-09-21Source: CoinGecko daily price series for ICP and BTC; drawdown method described belowNo specific drawdown percentage or recovery duration is asserted; the method and its limits are described.
Dataset, period and method
The dataset is the daily price series for ICP and BTC published by CoinGecko. The ICP period runs from the May 2021 listing to the review date; the Bitcoin period runs from the start of this site's daily record. The method is the standard one: a drawdown is the percentage decline from a running peak to a subsequent trough, and recovery time is the interval from that trough back to the level of the peak that preceded it.
Two conventions matter and are applied consistently. First, the running peak is the highest price observed up to that date, not the highest price in a fixed calendar window, so a drawdown is measured against the actual high-water mark. Second, recovery is measured from the trough rather than from the start of the decline, which is the convention this site uses on its recovery time page.
The comparison is therefore methodologically symmetric: the same arithmetic is applied to both series. What is not symmetric is the amount of history. Bitcoin's record contains several complete peak-to-trough-to-recovery sequences. ICP's record contains at most one, and it is still open.
What the two records can be compared on
| Measure | Bitcoin | ICP |
|---|---|---|
| Maximum drawdown depth | Measured across several completed cycles | Measured across one open sequence |
| Drawdown duration (peak to trough) | Multiple completed observations | One completed observation at most |
| Recovery time (trough to prior peak) | Several completed recoveries | Not yet established for the largest decline |
| Number of independent episodes | Enough to describe a range | Too few to describe a range |
Last reviewed 2026-09-21Source: CoinGecko daily series; drawdown convention as described aboveThe table describes what is measurable, not what the measurements are.
The asymmetry in the last row is the one that decides how the comparison should be read. Bitcoin's drawdown record can be summarised as a range — the deepest decline, the shallowest, the typical duration — because there are enough episodes to describe a distribution. ICP's record cannot be summarised that way, because a single observation has no distribution. Reporting "ICP's maximum drawdown is X" would be reporting one number from one episode and presenting it as a characteristic of the asset.
What can be said is directional and modest: ICP's traded history includes a decline of substantial depth, and the recovery from that decline has taken a long time relative to the length of the series. Both statements are about the observed record rather than about a tendency, and neither supports a claim about what ICP will do in a future decline.
Why the comparison is not like-for-like
The first reason is age. Bitcoin's record spans more than a decade and several market cycles; ICP's spans a single cycle and part of the next. A long series contains a wider variety of conditions, and a short one is more likely to be dominated by the particular conditions of its own period.
The second is the starting point. ICP's series begins at its listing, which is a period of price discovery rather than a steady state. The first months of a newly listed token are not representative of its later trading, and a drawdown measured from a listing-period peak is measured from an unusually volatile reference point.
The third is supply. ICP's circulating supply has changed substantially since listing, through the network's vesting and release schedule. A price series alone does not capture that, and a drawdown measured in price terms can differ from one measured in capitalisation terms. The market-capitalisation history page works through that effect directly.
The general point is covered on this site's drawdown explained page, which sets out why a drawdown is not the same thing as a loss and why the reference peak matters.
Limitations
The comparison rests on one complete ICP episode at most. Any statement about ICP's drawdown behaviour is therefore a statement about a single historical sequence, and single sequences are poor guides to future behaviour.
The series are venue-aggregated daily closes. Intraday extremes are not captured, so a drawdown measured on daily closes will understate the deepest intraday decline. The same limitation applies to both assets, so the comparison is at least consistent, but neither figure is the true maximum decline.
Finally, this page does not claim that ICP's drawdowns were caused by Bitcoin's. Two assets declining in the same period may be responding to a common market factor. The page on ICP during major BTC moves treats the timing question separately and explicitly as association rather than causation.
Sources and references
The price series come from CoinGecko. The drawdown and recovery conventions are this site's own and are applied identically to both assets.
- ICP historical price data. CoinGecko, Internet Computer (ICP): the daily series from the May 2021 listing.
- Bitcoin's drawdown record. Bitcoin Data Guide, Drawdowns: every major peak-to-trough decline in the daily record and how long recovery took.
- The drawdown convention. Bitcoin Data Guide, Drawdown Explained: why a drawdown is measured from a running peak and how it differs from a loss.
Related reading
- Research HubEvery dataset on the site, with methodology and provenance.
- Altcoin ResearchAltcoins measured against Bitcoin: design intent, consensus, execution, scaling and market structure.
- The ETH-BTC Correlation RecordHow the correlation is measured, how it behaves across windows, and where it breaks down.
- The ETH/BTC RatioWhat the ratio measures, how to read its trend, and why it is not a forecast.
- ETH During Bitcoin Bull PhasesAssociation within a common market factor, and what co-movement cannot establish.
- ETH During Bitcoin Bear PhasesDrawdown depth and duration compared over identical windows, and the limits of the comparison.