Did The Hedge Actually Work?

A trader puts on a hedge, the position moves against them, and the P&L still looks bad. Did the hedge fail — or did it do exactly its job and simply not eliminate all the pain, which is a different thing entirely? Without a way to measure hedge effectiveness after the fact, it's easy to draw the wrong conclusion in either direction: blaming a hedge that worked fine, or trusting one that quietly stopped working weeks ago.

What "worked" actually means

A hedge is not meant to make P&L flat. It's meant to reduce the variance of the combined position relative to the unhedged position. The standard way to check this is to compare how much the hedged position moved versus how much the underlying alone moved over the same period — if the combined position swung far less than the naked exposure would have, the hedge did real work, even if the combined P&L still isn't zero.

A common summary statistic is the hedge effectiveness ratio: one minus the variance of the hedged position's returns divided by the variance of the unhedged position's returns over the same window. A ratio near 1 means the hedge removed almost all the variance; a ratio near 0 means it removed almost none.

Worked example

Over a month, an unhedged $10m position in a stock has a return standard deviation that would have produced roughly $400,000 of P&L swing. With the hedge on, the actual combined position swung by only $120,000 over the same period. The variance reduction is substantial — the position that would have moved $400,000 moved $120,000 instead — telling the desk the hedge captured most of the relevant risk, even though the $120,000 residual (basis risk, timing gaps, imperfect ratio) is real and worth tracking separately.

What breaks the measurement

• Too short a window. A week of quiet markets tells you little about how a hedge performs in a real move.
• Comparing to the wrong counterfactual. You need to know what the position would have done unhedged, not just look at the hedged P&L in isolation.
• Ignoring costs. A hedge that reduces variance but bleeds premium or funding cost every day it's on isn't obviously "working" just because it caught one bad move — the ongoing cost has to be weighed against the protection.

What this means in practice

Desks that run hedges systematically usually review effectiveness on a regular cadence — monthly or after every large market move — rather than only when something feels wrong. A hedge that was effective at inception can degrade silently as correlations shift, so the measurement needs to be repeated, not done once and filed away.