Excess return: portfolio return above the risk-free rate or a benchmark
Excess return is the portfolio's return minus a defined reference rate. The reference is typically the risk-free rate (in which case excess return measures compensation for taking risk) or a benchmark index (in which case it measures active-management performance). The metric is the numerator of nearly every risk-adjusted return ratio—Sharpe, Sortino, Treynor, alpha—and the foundation of any meaningful return comparison.
What excess return is
Excess return is defined as Rₚ − Rᵣ, where Rₚ is the portfolio's return and Rᵣ is the chosen reference rate over the same period. The choice of reference matters: against the risk-free rate, the excess return is the compensation the investor earned for bearing risk; against a benchmark, it is the active return delivered by the manager or strategy relative to a passive alternative.
The two interpretations correspond to two distinct uses. Risk-adjusted return ratios use the risk-free rate as the reference because the question they answer is "how much return did the strategy earn per unit of risk taken?"—and the answer is meaningful only against the cost of taking no risk. Manager-evaluation ratios use a benchmark because the question is "how much value did the manager add over what a passive investor could have earned?"
The active-return interpretation is what the literature usually calls active return; excess return is the broader term that subsumes both meanings depending on context.
How it is used
The simplest use is direct comparison. A portfolio earning 8% in a year when T-bills yield 5% delivered 3% in excess return; the same portfolio in a year when T-bills yield 1% delivered 7%. The headline 8% looks the same in both years, but the contribution attributable to risk-bearing is meaningfully different.
Most risk-adjusted ratios divide excess return by a measure of risk. Sharpe = excess return / total volatility; Sortino = excess return / downside volatility; Treynor = excess return / beta. Each of these answers a slightly different question (per unit of total risk, downside risk, or systematic risk), but they share the same numerator: only the return earned above the reference is treated as compensable.
For benchmark-relative analysis, excess return forms the basis of the information ratio (excess return / tracking error) and active alpha (excess return adjusted for systematic exposures). Both metrics measure manager skill on the same underlying logic: skill is whatever return the manager produces above what a passive alternative would have delivered.
What the evidence shows
Long-run excess returns of major asset classes are well-documented. The equity risk premium—the long-run excess return on equities over Treasuries—has averaged approximately 4–6 percentage points per year across major markets over the post-1900 period (Dimson, Marsh, & Staunton). The fixed income excess return—long bonds over T-bills—has averaged approximately 1–2 percentage points over the same window. Each represents the long-run compensation for bearing the corresponding risk.
Manager-relative excess returns are more variable and more contested. The empirical literature on active equity management (Fama, French, Carhart, and many subsequent studies) has consistently found that the average active manager produces small or negative excess returns net of fees against an appropriate benchmark, with substantial dispersion across managers. The headline finding is that manager skill, when it exists, is concentrated in a minority of managers and is hard to identify ex ante.
The choice of reference matters more than is often acknowledged. An equity manager whose portfolio delivers 12% in a 10% market shows 2% excess return against the S&P 500—but if the portfolio's actual factor exposures (size, value, momentum) imply a 13% benchmark return, the genuine excess return is −1%. Active return without factor adjustment overstates skill in proportion to the manager's factor tilts.
Limitations and trade-offs
The reference rate's choice is meaningful. Comparing excess returns across periods requires the same reference convention; comparing across managers requires the same benchmark. Subtle differences in reference (T-bill vs cash deposit rate, total-return vs price-return benchmark, gross vs net of fees) can produce gaps of 0.5–1.0 percentage points in reported excess return without any change in underlying performance.
Excess return is also a univariate summary that loses information about the path. A strategy that earns 5% excess return through smooth quarterly contributions of 1.25% has the same excess return as one that produces 15% in one quarter and 0% in the others—but the two strategies are very different. The risk-adjusted ratios (Sharpe, Sortino) recover some of the path information by dividing by a risk measure, but excess return alone does not.
For multi-period comparisons, the convention of compounding excess returns versus arithmetic averaging matters. The geometric (compounded) excess return is what an investor actually experienced over a multi-period horizon; the arithmetic (averaged) excess return is the right input to forward-looking expected-return calculations. Both have valid uses; conflating them produces systematic error.
Excess return in pfolio
pfolio reports excess return implicitly through the Sharpe and Sortino ratios, both of which use return minus the configurable risk-free rate as their numerator. The risk-free rate setting and the resulting metrics are visible in pfolio Insights.
Related articles
Disclaimer
Get started now
