A benchmark portfolio is a specified reference portfolio used to compare an investment portfolio’s returns, holdings or risks. It defines the starting point for judging active decisions: a fund can make money while falling behind its benchmark, or lose money while beating it. The comparison is only useful when the benchmark suits the fund’s investment mandate.
A benchmark portfolio can be a hypothetical basket behind an index or a custom combination of asset-class portfolios. It need not be an actual funded account. Buying an index-tracking fund introduces expenses and implementation effects that the index calculation may omit.
First decide what the manager is responsible for
An equity manager restricted to one country should not automatically be judged against a global balanced portfolio. The comparison would mix stock selection with differences in asset class and geography.
For an asset owner, a policy benchmark often combines asset-class benchmarks using strategic asset-allocation weights. It represents the default allocation against which subsequent decisions are measured. The CFA Institute’s asset-owner guidance explains this distinction between the total fund’s policy reference and its underlying mandates.
Specify the reference before seeing the result. Record its constituents, weights and calculation rules, and establish why it fits the mandate. The GIPS guidance for firms discusses relevance, measurability and investability among benchmark-selection criteria. These are useful design principles; a familiar index name alone does not establish a fair comparison.
Build the return from opening weights
Consider an original synthetic benchmark starting 60% in equities and 40% in bonds. Over one holding period, their USD total returns are 10% and 2%. Assume reinvested income, consistent valuation times, no external cash flows, no trades during the period, and no fees, taxes or transaction costs. These are stipulated example conventions, not market quotes.
With opening weights summing to one, its portfolio return is
A USD 100,000 benchmark account would start with USD 60,000 in equities and USD 40,000 in bonds. Those sleeves become USD 66,000 and USD 40,800: USD 106,800 in total. This dollar ledger independently reconstructs the weighted-return calculation.
Original synthetic one-period USD ledger. Both accounts start with USD 100,000; returns include reinvested income and exclude flows, trades and costs.
The formula uses beginning-of-period weights. End weights already incorporate the returns being measured. For later periods, a buy-and-hold benchmark lets weights drift, while a benchmark with scheduled rebalancing resets them under stated rules. “60/40” alone does not specify a multi-period return series.
What does beating it by 1.2 mean?
Suppose the managed account earns 8% over the same period and under the same example conventions. Its arithmetic active return is
It ends at USD 108,000, USD 1,200 ahead of the benchmark. The proportional gain in ending wealth relative to the benchmark uses a different denominator:
Thus 1.2 percentage points of active return and about 1.12% more ending wealth describe different quantities. Label the convention when reporting outperformance.
One positive active return does not establish manager skill or risk-adjusted alpha. It may reflect risk exposures, allocation choices or luck. Nor does it measure tracking error, which concerns the variability of a series of active returns.
The benchmark also changes how positions are read
If is the portfolio’s opening weight and the benchmark’s, define its active weight as . When both weight vectors sum to one,
The return identity assumes the same asset-return vector, valuation period and frictionless holding convention for both portfolios. A negative active weight means an underweight; it can occur in a long-only portfolio without an actual short sale.
For the optimisation that follows from this decomposition, the Mean-Variance Optimisation lesson develops active weights and benchmark-relative risk in §2.3. The public page is a preview; the full lesson requires membership.
Before interpreting a real return gap, align the reporting period, currency and hedging treatment, income reinvestment, and fee/tax basis. Comparing a fund’s total return with a price-only index can create apparent outperformance from income alone. Custom benchmark weights and rebalancing rules also need explicit documentation; the GIPS firms guidance addresses both points for GIPS reporting.
Sources
- CFA Institute, Guidance Statement on Benchmarks for Firms, revised July 2023: pp. 1, 4, 7 and 21. Benchmark role, selection, custom blends and total-return comparisons.
- CFA Institute, Guidance Statement on Benchmarks for Asset Owners, April 2023: pp. 1–2. Policy benchmarks and underlying mandates.
- Edwin J. Elton, Martin J. Gruber, Stephen J. Brown and William N. Goetzmann, Modern Portfolio Theory and Investment Analysis, 9th edition, Wiley, 2014: Chapter 26, p. 692, and Chapter 28, p. 718. Benchmark choice, mandate-relative attribution and active positions relative to a passive portfolio.