The Calmar Ratio weighs what was earned against the deepest hole the asset fell into, leaving day-to-day choppiness out of it entirely. That answers the question anyone actually holding the thing wants answered: what did I have to sit through to collect this?

That suits assets which run hard and collapse hard. What hurts an owner is the hole, not the jitter, and this is the measure that grades performance against the part that hurt.

What it actually measures

The reading compares what the asset returned over a window with the deepest loss inside that same window. A high ratio means the return arrived without a punishing hole along the way; a low one means the return came at a cost.

The look-back setting is not cosmetic. Tighten it and the measure responds faster and jumps about more; widen it and it starts describing an era. Moving that control alters the question being asked, not merely the tidiness of the answer.

The line steps rather than drifts, because one day sets it

Most ratios move smoothly because everything feeding them is an average. This one does not. The worst loss it measures against is a single historical event, so the reading holds steady until a deeper one arrives and then resets in a single session.

The same mechanism ambushes readers going the other way. Once the deepest hole passes out of view entirely, the score can jump without anything at all happening in the present. Before treating a sudden improvement as fresh strength, check whether an old disaster has simply expired.

What it does not tell you

It rests entirely on one observation. A single unusual session governs the reading for as long as it stays in view, so the number is hostage to whether that day is still being counted.

The shape of the drawdown is invisible to it. A fast crash and a long grind that reach the same depth score identically here, and those are very different experiences to have lived through.

It is also blind above the peak. Everything the ratio knows about risk comes from declines, so an asset that has risen violently without ever falling far will read as low risk right up until it falls.

How to read it

Overbought. Return measured against the worst loss in the window sits in the top band of its own record, so the asset has been well paid for the pain it imposed.

Neutral. The ordinary band, where the return earned and the drawdown survived are broadly in their usual proportion.

Oversold. The bottom band: a deep loss to survive with little return to show for it.

The Calmar Ratio sits among the views on the Risk-Adjusted Returns dashboard, next to Risk of Ruin, Positive HODL Days and the Risk-Adjusted Composite.

Common questions

Why use Calmar for a volatile asset?

Because what hurts an owner is the hole, not the jitter. Grading returns against the deepest loss on record is closer to what sitting through the thing actually felt like.

Why does it jump in steps instead of drifting?

Because one past disaster sets the yardstick instead of an average across many. It sits still until something worse turns up and shifts it in a single session.

What happens when the worst drawdown drops out of view?

The score can jump without anything happening in the present, which is worth knowing before treating a sudden improvement as fresh strength.

How should the look-back be set?

It decides how much history each reading covers. Changing it changes what is being measured, not just how smooth the line is.

Does it capture the shape of a decline?

No. A fast crash and a long grind reaching the same depth score the same, and those are very different to sit through.