Research · Altcoins · Quant Research

Comparative risk-adjusted returns: BTC, ETH, SOL and ICP

A risk-adjusted return ratio divides a return by a measure of the risk taken to earn it. The ratio is only as meaningful as the two inputs, and for a panel with unequal histories the risk-free rate and the short sample matter more than the ranking. This page sets out the definitions and the caveats.

Last reviewed 2026-09-21Source: Methodology reference; no dataset is published on this pageRatio definitions follow standard practice; no ratio value is asserted.

The ratio definitions

The Sharpe ratio is the average excess return over the risk-free rate divided by the standard deviation of that excess return. It answers the question "how much return did this asset produce for each unit of volatility it exhibited". The Sortino ratio replaces the standard deviation in the denominator with the downside deviation, which counts only the returns below a target, usually zero or the risk-free rate. It answers a narrower question: how much return was produced for each unit of downside movement, ignoring upside volatility.

The distinction matters because the two ratios can rank assets differently. An asset with a few very large positive days has a high standard deviation and a high Sharpe denominator, which penalises it for upside movement that no holder experienced as risk. The Sortino ratio does not penalise that movement, so the same asset can rank higher on Sortino than on Sharpe. Neither ratio is more correct; they measure different things, and a comparison that reports only one has chosen a definition of risk without saying so.

A third measure, the Calmar ratio, divides the annualised return by the maximum drawdown. It is the most intuitive of the three for a long-horizon holder, because the denominator is the worst experience the asset produced rather than a statistical dispersion. It is also the least stable, because it rests on a single episode: change the window and the maximum drawdown can change substantially, which changes the ratio more than any change in the return.

The inputs that decide the answer

The risk-free rate is the first input and the most often glossed. The Sharpe ratio is defined on excess return, so the rate subtracted from the asset's return has to be stated and has to be the rate that applied over the window measured. A study that uses today's rate for a window that spans a period of near-zero rates overstates the excess return, and one that uses a single constant rate across a decade of changing policy does the same. The rate is a choice, and it moves the numerator directly.

The return definition is the second input. An arithmetic average of periodic returns and a geometric average produce different figures, and the geometric average is the one that corresponds to what a holder actually experienced, because it accounts for compounding. A study that reports an arithmetic mean return in a Sharpe ratio is reporting a number no holder earned. The annualisation convention is the third: scaling a monthly ratio to a yearly one by multiplying by the square root of twelve assumes the returns are independent, which they are not, and it inflates the figure.

The window is the fourth and the most consequential. A ratio computed over a bull market will be high for every asset in the panel, because the numerator is large and the denominator is measured over a period of rising prices. A ratio computed over a window containing a crash will be low or negative for every asset. The ranking between assets is more stable than the level, but it is not immune: an asset whose history begins after the crash in the window will have a denominator that excludes it.

The short-history caveat for ICP

The Internet Computer's public price history begins in May 2021, which makes it the shortest series in the panel by a wide margin. It is shorter than Solana's by about a year, shorter than Ethereum's by about six years, and shorter than Bitcoin's by more than a decade. That has three consequences for a risk-adjusted comparison, and all three push in the same direction.

First, the sample is small. A ratio estimated from a few years of daily data has a wide confidence interval, and the interval is wider for the Sortino and Calmar ratios than for the Sharpe because they depend on a subset of the observations. Second, the period is not representative. The window beginning in May 2021 contains the decline from the 2021 peak and the recovery that followed, and it contains no full cycle of its own. A ratio measured over that period describes that period, and there is no longer history against which to check whether it is typical.

Third, the comparison is structurally unfair in both directions. The Internet Computer's series excludes the 2017 and 2013 episodes that dominate Bitcoin's maximum drawdown, which flatters it on a Calmar ratio. It also excludes the early, high-growth years of Bitcoin's history, which flatters Bitcoin on a Sharpe ratio. Neither effect is a property of the assets; both are properties of when their public price series began. A comparison that does not say so is comparing four different periods and calling it a ranking.

Method, dataset and limitations

The ratio definitions a risk-adjusted comparison has to state, and what each choice changes about the result.
Ratio	Numerator	Denominator	Principal caveat
Sharpe	Average excess return over the risk-free rate	Standard deviation of the excess return	Penalises upside volatility as though it were risk, and depends on the risk-free rate chosen.
Sortino	Average excess return over the target	Downside deviation below the target	Depends on the target chosen and on a subset of observations, so it is noisier than Sharpe on a short sample.
Calmar	Annualised return	Maximum drawdown over the window	Rests on a single episode; changing the window can change the ratio more than any change in the return.

Last reviewed 2026-09-21Source: Methodology reference; standard performance-measurement practiceNo dataset is published on this page and no ratio value is asserted.

The dataset is the same panel of daily closes described on the volatility page, with the risk-free rate drawn from a published short-term government series and stated for the window measured. The period is bounded by the shortest history in the panel for a common-window comparison, or stated per-asset for a full-history comparison. The retrieval date is recorded for both the price series and the rate series.

The limitations are the ones above: a risk-free rate that is a choice, a return definition that is a choice, an annualisation convention that assumes independence, a window that decides the level, and a panel whose shortest member has a history of a few years. A risk-adjusted comparison that states all of that is a description of four different periods. One that reports a ranking without them is a ranking of when each series began.

Sources and references

The ratio definitions follow standard performance-measurement practice. The series a study would use are named and linked below.

Daily price series for the four assets. CoinGecko, CoinGecko API documentation: the historical daily close series used to build the return panel, with the retrieval date recorded on each study.
Reference and cross-check series. Coin Metrics, Community Network Data: an independent daily series used to check that a price move is not an artefact of a single venue's quote.
On-chain and market-structure context. Glassnode, Glassnode API documentation: the realised-capitalisation and supply series used to describe the market each asset trades in, not to compute the return panel.
Asset-level reference data. Messari, Messari API documentation: a third series used where the first two disagree, so the disagreement can be reported rather than hidden.
Bitcoin's own record, for the baseline. Bitcoin Data Guide, Bitcoin Price History and Data Sources & Methodology: the site's own compiled daily record and the provenance rules that apply to it.

The Sharpe ratio. William Sharpe, The Sharpe Ratio: the definition of the ratio and the excess-return convention it assumes.
The Sortino ratio and downside deviation. Sortino and Price, Performance Measurement in a Downside Risk Framework: the case for measuring risk on the downside only, and the target the denominator depends on.
Bitcoin's own risk-adjusted record. Bitcoin Data Guide, Risk-Adjusted Returns and CAGR Explained: the ratio definitions used across this site and why a compound growth figure hides the path a holder lived.
The risk-free rate series. U.S. Department of the Treasury, Daily Treasury Bill Rates: the published short-term rate series a study would subtract from the asset return.

Last reviewed 2026-09-21. No live market data is fetched or displayed on this page.