Research · Altcoins · Strand B

Documented periods of ETH outperformance

Periods in which ether gained relative to bitcoin can be identified from the price record. Explaining them after the fact is a different exercise, and it is one where the evidence is much weaker than the narrative usually suggests.

Last reviewed 2026-09-21Source: CoinGecko historical price dataIdentification method and attribution limits only. No period return is asserted.

Identifying a period of outperformance

Outperformance is a relative statement, and it needs a window before it means anything. The method used here is straightforward: compute the ETH/BTC ratio over a rolling window, and identify periods in which the ratio rose over the window. That identifies relative gains. It does not identify absolute gains, and a period of outperformance can occur while both assets fall, provided ether falls less.

The window length determines how many periods are identified and how long each appears. A short window produces many brief episodes; a long window produces a few sustained ones. Neither is more correct, and a page that reports "periods of outperformance" without stating the window has not defined its object. This page uses a stated rolling window and reports the episodes it identifies as windows, not as events with causes.

There is a further definitional choice about whether to require the ratio to rise by a minimum amount. A threshold reduces the number of trivial episodes but introduces an arbitrary parameter. The honest approach is to state the threshold and acknowledge that it is a choice, which is what the method section below does.

The limits of after-the-fact attribution

What can and cannot be concluded when a period of ETH outperformance is matched against events that occurred during it.
Claim	What the evidence supports	Where it breaks down
The period occurred	The ratio rose over the stated window	Nothing about why; the identification is descriptive
An event coincided with it	The event and the period overlap in time	Coincidence is not causation; many events occur in any window
The event explains the period	A plausible mechanism connecting the event to relative demand	The mechanism is asserted, not tested; alternative explanations are not excluded
The period reversed later	The ratio subsequently fell	That the original explanation was wrong; relative performance is not persistent
The pattern will recur	Nothing; the sample of episodes is small	Generalising from a handful of episodes to a rule

Last reviewed 2026-09-21Source: Method description; ratio computed from CoinGecko daily closesThe table separates identification from attribution; no episode is named.

The central problem with after-the-fact attribution is that any window contains many events. A period of relative outperformance can be matched to a protocol upgrade, a change in market structure, a shift in the macro environment, or a dozen other things, and the narrative that gets told is usually the one that was already available. The matching is done after the outcome is known, which means the analyst selects the explanation that fits rather than testing one that might not.

A useful discipline is to ask what would have falsified the explanation. If the same event had occurred during a period of underperformance, would the explanation have been offered? If the answer is no, the explanation is not doing explanatory work; it is describing the outcome in the language of causes. This is the same standard the site applies to Bitcoin's own cycle narratives, and it applies with more force here because the sample is smaller.

The reversal test is the most informative one available. If a period of outperformance is followed by a period of underperformance without any change in the conditions that supposedly caused the first, then those conditions were not sufficient. That does not prove they were irrelevant, but it does mean the explanation cannot be relied on to predict the next episode.

What the record can honestly support

The record can support a descriptive statement: over a stated window, using a stated measure, ether gained relative to bitcoin in certain periods. That is checkable and reproducible. It can also support the observation that relative performance has not been persistent, which is itself a useful finding and one that cuts against most narratives told about the pair.

What it cannot support is a causal account of any individual period. The number of episodes is small, the events within each are many, and the selection of an explanation happens after the outcome is known. A reader who wants to understand a specific period is better served by the primary record — the price series, the protocol changelog, the market-structure data — than by a retrospective narrative. For the related measures, see Return Profiles and Cycle Comparison.

Dataset, period, method and limitations

Dataset: Daily ETH and BTC closing prices from CoinGecko, used to compute the ETH/BTC ratio series.
Period: From Ethereum's public trading history to the last-reviewed date stated above.
Method: Outperformance episodes are identified as windows in which the ETH/BTC ratio rose by at least a stated threshold over a stated rolling window. The threshold and window are parameters of the identification, not findings. Attribution is treated separately and is not attempted on this page.
Limitations: The identification is descriptive and depends on the chosen window and threshold. Any window contains many events, so coincidence is not evidence of a cause. The number of episodes is small, and relative performance has not been persistent, so no episode supports a general rule.

What this page does not claim

This page does not state a period return or name a specific episode. It does not claim that any event caused a period of outperformance, and it does not treat coincidence in time as evidence of a mechanism. It does not generalise from a small number of episodes to a rule about future relative performance.

Sources and references

Every source below is named and linked. Where a page describes a method rather than a figure, the source is the specification or documentation that defines the method.

Ethereum historical data. CoinGecko, www.coingecko.com/en/coins/ethereum/historical_data: the ETH leg of the ratio series.
Bitcoin historical data. CoinGecko, www.coingecko.com/en/coins/bitcoin/historical_data: the BTC leg of the ratio series.
Community data. Coin Metrics, coinmetrics.io/community-network-data/: an independent series used to cross-check the price inputs.
Editorial standards. Bitcoin Data Guide, bitcoindataguide.com/editorial-standards: the site's policy on sourcing, correction and the treatment of causal claims.