Upside and Downside Capture Ratios

A single beta number describes how a portfolio moves on average relative to its benchmark, treating a rising market and a falling one as symmetric. Real managers usually aren't symmetric — a defensive fund might participate in only 60% of a rally while cushioning 80% of a selloff, and beta alone can't show that split. Capture ratios separate the two regimes on purpose, answering "how much of the good months did you get" and "how much of the bad months did you avoid" as two different questions.

Two averages, in two different regimes

$$\text{Upside capture} = \frac{\overline{r}_{\text{portfolio, benchmark up}}}{\overline{r}_{\text{benchmark, up}}}, \qquad \text{Downside capture} = \frac{\overline{r}_{\text{portfolio, benchmark down}}}{\overline{r}_{\text{benchmark, down}}} .$$

In words: split the historical months into two buckets — months the benchmark was up, and months it was down. In the up bucket, average the portfolio's return and divide by the average benchmark return in that same bucket; that's upside capture, expressed as a percentage. Do the same in the down bucket for downside capture. A capture ratio above 100% means the portfolio moved more than the benchmark in that regime; below 100% means it moved less.

Each point on a scatter like this is one month's benchmark return (horizontal) against the portfolio's return (vertical). Capture ratios are just the average slope of the points sitting to the right of zero (up months) versus the average slope of the points sitting to the left of zero (down months) — two different regressions instead of one.

Worked example 1 — computing both ratios

Over ten months, the benchmark was up in six of them, averaging $+3.0\%$ in those months; the portfolio averaged $+2.4\%$ in the same six months. Upside capture: $2.4/3.0 = 80\%$. The benchmark was down in the other four months, averaging $-2.5\%$; the portfolio averaged $-1.5\%$ in those same four months. Downside capture: $-1.5/-2.5 = 60\%$. This fund captured 80% of the market's gains but only 60% of its losses — a genuinely defensive profile, not just a lower-beta one, since the two numbers differ from each other.

Worked example 2 — same beta, different story

A second fund has almost the same overall beta as the first, roughly 0.75, but its capture ratios tell a different story: upside capture of $70\%$ and downside capture of $75\%$ — it participates less in rallies than it does in selloffs. Two funds with nearly identical single-number betas can have opposite-shaped participation once split into up and down regimes, and only the split reveals which one is the more defensively built portfolio.

What this means in practice

Allocators use capture ratios to screen for managers whose downside protection isn't just an artifact of low overall volatility — a fund with upside capture meaningfully above its downside capture is adding value asymmetrically, which a single beta or standard deviation figure would hide entirely. It's a standard slide in any long-only manager's pitch deck for exactly this reason.