What it shows

Every coin in the top 300, plotted by the risk it carried and what it paid for carrying it. Volatility runs along the bottom, annual return up the side. Each bubble is sized by market value and coloured by its Sharpe ratio — green where the return justified the risk, red where it did not.

The two dashed lines are the ones to read first. Above the Sharpe 1 line, a coin returned more than one unit of gain for each unit of volatility. Above Sharpe 2 is exceptional, and worth checking it is not a single spike. Below the dotted line, the coin lost money.

How to use it

The useful direction is up and to the left: return without much turbulence. A coin high on the right made money but put you through a great deal to get it, which matters if you would have sold on the way.

Use the buttons above the chart to filter by size. The tiers behave very differently, and that is where this becomes personal to how you invest.

A large bubble sitting high and to the left is a big, calm, profitable coin — the rarest thing on the chart. Small bubbles far to the right are the opposite: tiny, wild, and usually below breakeven.

Moving through history

Drag the slider to step back through every weekly snapshot to October 2024. Watch the cloud: in a broad rally the whole thing lifts, while in a rotation it stretches, some coins climbing as the rest sink. A single snapshot cannot tell you which of the two you are in.

Both axes and the colours stay fixed, so the movement you see is real. Bubble sizes are relative to each date — compare them within a snapshot, not across.

The fine print

The steps are evenly spaced but the dates are not. Nothing ran between April and August 2026, so one step there covers five months.

Return and volatility are measured over the trailing year at each weekly snapshot. The Sharpe ratio here is simply return divided by volatility, with no risk-free rate subtracted — which is why a given Sharpe is a straight line from the origin, and why the reference lines can be drawn at all.

Volatility was only recorded directly from August 2026. Earlier dates recover it as return divided by Sharpe, which is exact arithmetic rather than an estimate. The few coins whose return was exactly zero cannot be recovered that way and are absent from those dates.

Coins under 15% annual volatility are left out entirely. They are stablecoins, wrapped assets and tokenised funds, which are not observations about risk and reward. In the latest snapshot that excludes exactly one coin.

The return axis stops at −200% and +500%. Coins beyond it are pinned to the edge as triangles with their true figure in the hover — plotting one coin at +32,000% to scale would squash the other 299 into a single dot. Volatility is on a logarithmic scale because it spans nearly 300 times, from 15% to over 4,000%.

A coin with very little price history can produce a flattering Sharpe from very few real observations. Treat an unfamiliar name in the top corner as a question rather than an answer, and check it in the screener file before acting on it.