Guide

Upside and downside capture ratio explained

Harbor Capital's 2024 manager diligence cycle reviewed a large-cap growth fund whose five-year annualized return beat the Russell 1000 Growth index by 80 basis points. On paper, the mandate looked healthy. Then the performance team computed upside and downside capture ratios — metrics that ask a different question than cumulative return: when the benchmark rose, how much of that rise did the fund capture, and when the benchmark fell, how much of that fall did the fund share? Upside capture was 112% (the fund gained 12% more than the index on up months) but downside capture was 118% (it lost 18% more on down months). The asymmetry meant investors experienced a bumpier ride with no net benefit after fees. Trustees flagged the mandate for reduction. Capture ratios translate abstract beta and volatility into intuitive bull/bear behavior. This guide defines upside and downside capture, walks through the formulas and period definitions, explains capture ratio pairing and batting average, compares capture metrics to Sharpe, Sortino, and Jensen's alpha, works a Harbor Capital sleeve example, provides a metric decision table, lists pitfalls, and ends with an allocator checklist.

What capture ratios measure

Most performance metrics collapse an entire return history into one number. Capture ratios split the sample by benchmark direction: months (or quarters) when the benchmark had a positive return versus months when it had a negative return. They answer whether a manager is a “good loser” (downside capture below 100%) and a “good winner” (upside capture at or above 100%) — or the opposite.

A passive index fund should show roughly 100% upside and 100% downside capture vs its own index (minus fees and tracking error). Active managers deliberately deviate: a defensive equity fund might target 85% downside capture and accept 90% upside capture. A high-conviction growth fund might accept 120% downside to chase 130% upside. The pair tells you the manager's stated style matches realized behavior.

Upside capture ratio

Over all periods where the benchmark return R_b > 0:

Upside Capture = [Σ R_p] / [Σ R_b] × 100%

Where the sums run only over up periods and R_p is the fund return in those same periods. Some data vendors compound returns within the up bucket instead of summing; the concept is identical: fund participation in benchmark gains. A value above 100% means the fund captured more than the benchmark's upside; below 100% means it participated less.

Downside capture ratio

Over all periods where the benchmark return R_b < 0:

Downside Capture = [Σ R_p] / [Σ R_b] × 100%

Because both numerator and denominator are negative in down periods, the ratio is positive. A downside capture of 80% means the fund lost only 80% as much as the benchmark during drawdowns — desirable for capital preservation. Above 100% means the fund amplified benchmark losses.

Capture ratio (pairing)

Analysts often report the pair as upside/downside, e.g. 95/85 (participates in 95% of gains, 85% of losses) or 110/105 (aggressive participation both ways). The ideal asymmetry depends on mandate: defensive allocators want upside/downside above 1.0 when expressed as a ratio of the two percentages (e.g. 95/85 = 1.12) or simply upside > downside in percentage terms for asymmetric participation favoring the investor.

How to compute capture ratios correctly

Small methodological choices shift capture ratios materially. Document these before comparing funds or vendors.

Period frequency and definition

Monthly returns are standard for equity mutual funds and ETFs; daily data suits hedge funds with rapid trading. Use the same frequency for fund and benchmark. Define “up” and “down” strictly by benchmark sign; zero-return months are typically excluded from both buckets.

Lookback window

Three years is a common minimum; five and ten years show regime behavior across bull and bear cycles. A fund launched in 2020 may show excellent downside capture simply because it never lived through a prolonged bear market — always pair point-in-time capture with maximum drawdown and full-cycle cumulative return.

Gross vs net returns

Capture ratios on net-of-fee returns reflect what investors experience. Gross capture flatters active managers; institutional due diligence usually computes both. Include reinvested dividends in total return series.

Benchmark alignment

Capture is meaningless against the wrong benchmark. A mid-cap fund measured vs the S&P 500 will show distorted capture because style drift dominates. Use the prospectus benchmark or a custom blended index for multi-asset funds.

Batting average

Related but distinct: batting average is the percentage of periods the fund beat the benchmark outright, regardless of magnitude. A manager can have a 55% batting average but poor capture if wins are small and losses are large. Use batting average alongside capture for a fuller picture of consistency vs magnitude.

Interpreting capture profiles

Capture ratios describe behavior, not skill by themselves. Context from mandate and market regime is essential.

100 / 100 — Index-like participation; active fees are hard to justify unless other services add value.
110 / 90 — Classic asymmetric active equity profile: participates more in rallies, less in selloffs. Attractive if persistent.
95 / 75 — Defensive profile; may lag in strong bull markets but protects in corrections.
120 / 130 — Amplifies both directions; high beta or concentrated bets. Requires strong risk tolerance.
80 / 110 — “Worse on both sides” or poor timing; red flag unless mandate is capital-preservation with hedges not reflected in the benchmark.

Rolling 36-month capture charts reveal style drift: a manager who marketed as defensive but whose downside capture crept from 85% to 115% over two years has changed risk posture without disclosure. Harbor Capital plots rolling capture on every active equity review slide.

Harbor Capital worked example

Harbor's investment committee evaluated two large-cap candidates for a $40M sleeve in Q1 2024. Both beat the Russell 1000 over five years.

Fund A (concentrated growth): five-year upside capture 114%, downside capture 109%. Net effect: slightly more participation in both directions. Cumulative outperformance came from a handful of big up months; batting average was only 52%. After 75 bps fees, net upside/downside was roughly 113/110 — modest asymmetry, not enough to justify active fees vs a low-cost growth ETF.

Fund B (quality bias): upside capture 96%, downside capture 78%. The fund gave up 4% of benchmark rallies but absorbed only 78% of benchmark declines. Over the 2018 Q4 and 2022 drawdowns, Fund B's absolute losses were 22% smaller than the index while participating in most of the 2019–2021 bull run. Harbor allocated to Fund B, pairing capture review with information ratio vs the Russell 1000 and a three-factor alpha check.

The lesson: headline outperformance hid Fund A's weak asymmetry; capture ratios surfaced Fund B's defensive value proposition in a form trustees could grasp without reading a beta regression.

Metric decision table

Question	Best metric	Why not capture alone?
How much benchmark upside/downside did we participate in?	Upside / downside capture	—
Return per unit of total volatility?	Sharpe ratio	Capture ignores magnitude of sideways periods and vol outside up/down buckets
Return per unit of downside volatility?	Sortino ratio	Sortino uses continuous downside deviation, not discrete down months
Skill after adjusting for market beta?	Jensen's alpha	Capture does not isolate systematic risk in one regression framework
Worst peak-to-trough loss?	Maximum drawdown	Capture sums monthly down periods; may miss intra-month path
Active return per tracking error?	Information ratio	Capture is directional; IR measures consistency of active bets
% of periods beating benchmark?	Batting average	Ignores win/loss magnitude

Common pitfalls

Short sample with few down months. Three years of bull market can show 70% downside capture from two lucky months — not a structural edge.
Wrong benchmark. Capture vs the S&P 500 for a small-cap fund measures style mismatch, not manager skill.
Summing vs compounding. Vendors differ; do not compare Morningstar capture to a custom sum-based spreadsheet without reconciling.
Ignoring fees. Gross upside capture overstates what LPs keep.
Conflating capture with alpha. 90/80 capture can come from low beta rather than stock-picking skill.
Survivorship bias. Fund databases drop dead funds that often showed 130% downside capture before liquidation.
Options overlay blind spots. Written puts can flatten downside capture in calm periods then fail catastrophically — pair with tail risk metrics.
Single-manager hero periods. One PM's capture profile may not survive team turnover; use rolling windows.

Allocator checklist

Compute upside and downside capture over at least 60 months of monthly total returns.
Use the fund's stated prospectus benchmark, not a convenience index.
Report net-of-fee capture for LP-facing materials; gross for manager evaluation.
Pair capture ratios with batting average and cumulative active return.
Plot rolling 36-month capture to detect style drift.
Compare capture profile to the manager's marketed style (defensive, benchmark-aware, high conviction).
Cross-check with maximum drawdown and Sortino for tail behavior.
Run Jensen's alpha or multi-factor regression to separate beta effects from selection.
Document period definition (monthly vs daily) and vendor methodology.
Re-audit capture after benchmark changes, fee revisions, or portfolio manager turnover.

Key takeaways

Upside capture measures fund participation in benchmark gains; downside capture measures participation in benchmark losses.
Below 100% downside capture is desirable for capital preservation; above 100% upside may be desirable for aggressive mandates.
The upside/downside pair reveals asymmetry that cumulative return and beta alone can hide.
Require long samples with multiple drawdowns; short bull-market windows mislead.
Combine capture with Sharpe, alpha, drawdown, and information ratio for complete manager due diligence.