The risk-free rate explained: what it is and why it shapes every return calculation

The risk-free rate is the return an investor can earn on a hypothetical investment with no default risk and no exposure to market fluctuations. It is the foundation of nearly every measure of risk-adjusted return—Sharpe, Sortino, alpha—and the benchmark against which every risky investment is, implicitly, evaluated.

What the risk-free rate is

In theory, the risk-free rate is the yield on an investment that delivers a known cash flow at a known future date with certainty. No real-world instrument fully satisfies this definition: even sovereign debt of high-quality issuers carries some default risk, and any nominal instrument carries inflation risk. In practice, the risk-free rate is approximated by short-dated government bills issued by issuers considered effectively default-free at the holding horizon—typically the three-month US Treasury bill, the three-month German Bund, or the equivalent in other major markets.

The choice of approximation matters. A short-dated US T-bill yield is the conventional risk-free rate for US dollar calculations and is a defensible proxy for an investor whose liabilities and consumption are denominated in USD. For a Swiss-franc-based investor, the same instrument carries currency risk and is no longer risk-free at the relevant horizon; the appropriate proxy is a short-dated CHF instrument. The risk-free rate is always a function of the currency, the horizon, and the issuer combined.

How it is used

The risk-free rate enters most return metrics as an offset. The Sharpe ratio, defined as (portfolio return − risk-free rate) / portfolio volatility, measures the excess return per unit of risk taken—meaning the return earned above what the investor could have obtained without taking risk at all. The Sortino ratio uses the same numerator with a different denominator (downside volatility). Alpha, in its simplest form, is the portfolio's return minus the risk-free rate minus the market beta times the market's excess return.

The economic logic is consistent across these metrics: only the return earned above the risk-free rate is attributable to the risk-bearing decision. A portfolio that earns 7% in an environment where T-bills yield 5% has delivered 2% of risk-bearing return; the same portfolio in an environment where T-bills yield 1% has delivered 6%. The headline 7% looks the same, but the contribution of risk-bearing differs by a factor of three.

The same rate also enters discounted-cash-flow valuation as the floor on required returns: any risky asset must offer an expected return above the risk-free rate, with the gap compensating for the risk borne. Without that anchor, risk-adjusted comparisons become arbitrary.

What the evidence shows

The level of the risk-free rate has changed substantially over time. US T-bill yields averaged approximately 4% over the post-war period through the late 1970s, rose to approximately 14% during the 1981 inflation peak, and fell to nearly 0% in the post-2008 zero-rate environment. The same Sharpe ratio computed for an equity portfolio over different decades will look meaningfully different depending on which risk-free rate level was prevailing—and on whether the metric uses a contemporaneous risk-free rate or a single fixed value across the whole period.

For long-run performance comparisons, the appropriate practice is to use the contemporaneous risk-free rate in each period rather than a single retroactive figure. This avoids the artefact of penalising or rewarding a strategy because it operated during a high-rate or low-rate regime. The Sharpe ratio of an equity strategy over 1995–2005 is not directly comparable to its Sharpe ratio over 2010–2020 unless both are computed against contemporaneous risk-free rates.

Within a given period, the choice between using a contemporaneous monthly T-bill yield versus a single annualised figure for the year typically produces small differences in the resulting metrics—at the 0.05 to 0.1 level for Sharpe ratios, depending on rate volatility. The choice still matters when comparing similar strategies, and conventions should be applied consistently across the comparison.

Limitations and trade-offs

The risk-free rate is a useful abstraction but not a real instrument. T-bills carry inflation risk; the yields are nominal; the implicit assumption that an investor can earn the risk-free rate without friction does not survive contact with transaction costs, custody fees, or imperfect rollover mechanics in practice.

The choice of horizon also matters. A three-month T-bill yield is the right risk-free rate for evaluating a short-horizon strategy; a long-horizon investor might reasonably argue that the appropriate comparison is to a long-dated TIPS yield (a real-return contract) rather than a short-dated nominal yield. Different horizons produce different comparisons, and the choice should match the investment problem rather than be inherited by convention.

The risk-free rate in pfolio

The risk-free rate used in Sharpe and Sortino ratio calculations is configurable in pfolio. The default reflects a representative short-term government bond yield; investors can override this to match their preferred reference rate. The setting and its effect on metrics are documented in the metrics we use.

Related articles

• Sharpe ratio explained: measuring risk-adjusted portfolio returns
• Sortino ratio: a better measure of downside risk-adjusted return
• Alpha in investing: what it means and how systematic strategies generate it
• Inflation-linked bonds: how TIPS and linkers protect purchasing power
• Excess return: portfolio return above the risk-free rate or a benchmark
• Money market funds and cash equivalents: the cash leg of a multi-asset portfolio