Third-Party Research & Methodology Only

This section shares summaries of third-party academic research and descriptions of quantitative models. The content represents the findings of the original researchers, not the opinions or recommendations of Foxholm Financial. Foxholm Financial does not publish hypothetical or backtested performance metrics on its quantitative research pages. All content is restricted to methodology, signal construction, factor logic, and risk architecture. SEC rules require that investment advisers not present misleading performance data, and our methodology-only approach reflects that standard and the firm's fiduciary obligations.

Risk-Adjusted Return Metrics

Concept Performance Measurement Risk Analysis

Robert Stowe, AAMS® Investment Advisor

Raw returns tell only half the story. An investment that gained 15% sounds impressive until you learn that its price swung wildly along the way, exposing the holder to severe potential losses. Risk-adjusted return metrics solve this problem by measuring how much return an investment generates relative to the risk it takes. The most widely used measures are the Sharpe ratio, the Sortino ratio, and the Calmar ratio, each capturing a different dimension of risk.

These metrics matter because they allow meaningful comparisons between investments with very different risk profiles. A low-volatility bond fund and a high-growth equity strategy might both return 10% in a given year, but they achieved that return with vastly different levels of uncertainty. Risk-adjusted metrics provide a common yardstick for evaluating whether the risk taken was adequately compensated.

Conceptual Framework

Every risk-adjusted metric follows the same basic logic: divide some measure of return by some measure of risk. The differences between metrics come from how each one defines "return" and "risk." The choice of metric matters because different definitions of risk capture different aspects of the investment experience.

The Sharpe Ratio

The Sharpe ratio, introduced by William Sharpe in 1966, is the most widely used risk-adjusted performance measure. It calculates the excess return (the return above the risk-free rate) per unit of total volatility (measured by standard deviation).

Formula: Sharpe Ratio = (Portfolio Return − Risk-Free Rate) ÷ Standard Deviation of Portfolio Returns

A Sharpe ratio of 1.0 means the portfolio earned one unit of excess return for each unit of volatility. A ratio above 1.0 is generally considered good, and above 2.0 is exceptional. A negative Sharpe ratio means the portfolio underperformed the risk-free rate.

The Sharpe ratio's strength is its simplicity and universality. It can compare any two investments regardless of asset class, strategy, or time horizon. Its weakness is that it treats all volatility equally: upside volatility (large gains) and downside volatility (large losses) both increase the standard deviation and reduce the ratio. An investment that occasionally produces large positive surprises is penalized the same way as one that occasionally produces large negative surprises.

The Sortino Ratio

The Sortino ratio, developed by Frank Sortino and Robert van der Meer in 1991, addresses the Sharpe ratio's key limitation by measuring only downside risk. Instead of using standard deviation (which captures all volatility), the Sortino ratio uses downside deviation, which counts only returns that fall below a minimum acceptable return (often set to zero or the risk-free rate).

Formula: Sortino Ratio = (Portfolio Return − Minimum Acceptable Return) ÷ Downside Deviation

The practical difference matters most for investments with asymmetric return distributions. A strategy that earns steady small gains with occasional large gains will have a higher Sortino ratio than Sharpe ratio because the large gains increase standard deviation but not downside deviation. Conversely, a strategy with frequent small gains and occasional large losses will have a lower Sortino ratio than Sharpe ratio because the losses dominate the downside deviation calculation.

Many practitioners prefer the Sortino ratio because investors do not actually experience upside volatility as risk. Unexpected gains are welcome; unexpected losses are the real concern. The Sortino ratio aligns more closely with how most people think about investment risk.

The Calmar Ratio

The Calmar ratio, named by Terry Young in 1991 in Futures magazine, measures the average annual return relative to the maximum drawdown (the largest peak-to-trough decline) over a specified period, typically three years.

Formula: Calmar Ratio = Annualized Return ÷ Maximum Drawdown

Where the Sharpe ratio uses volatility and the Sortino ratio uses downside deviation, the Calmar ratio uses the worst-case actual loss experience. This makes it particularly useful for evaluating strategies where the primary concern is surviving large drawdowns. A Calmar ratio of 2.0 means the strategy earned twice its worst decline in annualized returns. A ratio below 1.0 means the worst drawdown exceeded the annual return.

The Calmar ratio is especially relevant for retirement portfolios and strategies that cannot tolerate deep losses. A retiree drawing income from a portfolio cares more about the depth of the worst decline than about day-to-day volatility, making the Calmar ratio a more relevant metric than the Sharpe ratio for this context.

Other Risk-Adjusted Metrics

Several additional metrics complement the three primary ratios:

Treynor ratio: Similar to the Sharpe ratio, but uses beta (sensitivity to market movements) instead of standard deviation as the risk measure. The Treynor ratio measures excess return per unit of systematic risk rather than total risk. It is most useful when evaluating a portfolio that is part of a larger, diversified allocation, since diversifiable risk is not relevant in that context.
Information ratio: Measures the excess return relative to a benchmark (called active return) divided by the variability of that excess return (called tracking error). An information ratio of 0.5 or above is considered strong for active managers. This metric is the standard for evaluating managers who are trying to beat a specific benchmark rather than generate absolute returns.
Omega ratio: Compares the probability-weighted gains above a threshold to the probability-weighted losses below it. Unlike the Sharpe and Sortino ratios, the Omega ratio considers the entire return distribution, not just its first two moments (mean and variance). This makes it better suited for strategies with non-normal return distributions.

Risk Architecture

Choosing the Right Metric

The choice of risk-adjusted metric should match the investment context and the investor's primary concern:

General portfolio comparison: The Sharpe ratio remains the default because it is widely understood, easy to calculate, and applicable across all asset classes. When comparing two mutual funds, two ETFs, or two allocation strategies, the Sharpe ratio provides a reasonable starting point.
Downside-sensitive investors: The Sortino ratio is more appropriate when the investor cares primarily about avoiding losses rather than minimizing overall volatility. Retirees, endowments with spending requirements, and investors with short time horizons often fall into this category.
Drawdown-focused evaluation: The Calmar ratio is the right choice when the worst-case scenario matters most. Strategies that use leverage, concentrated positions, or illiquid assets should be evaluated with drawdown-based metrics because their maximum loss can be far more severe than their day-to-day volatility suggests.
Benchmark-relative evaluation: The information ratio is the standard for evaluating active managers against their stated benchmark. It answers the question: is the manager's deviation from the benchmark generating enough excess return to justify the additional variability?

Known Limitations

Limitations to Consider

All metrics are backward-looking: Risk-adjusted ratios are calculated from historical data and describe what has already happened. A high Sharpe ratio over the past five years does not guarantee the same ratio going forward. Market conditions change, strategies experience signal decay, and the risk profile of an investment can shift over time.
Normal distribution assumption: The Sharpe ratio assumes returns are normally distributed (bell curve shaped). In practice, financial returns exhibit fat tails (extreme events occur more often than a bell curve predicts) and skewness (the distribution is not symmetric). When returns are not normally distributed, the Sharpe ratio can understate the true risk. The Sortino ratio partially addresses this by focusing on downside risk, but it still relies on a summary statistic rather than the full distribution.
Sensitivity to the measurement period: The same strategy can produce very different risk-adjusted ratios depending on the time period selected. Including or excluding a single bad quarter can change a Calmar ratio dramatically. Comparing metrics across different time periods is misleading.
Risk-free rate dependency: The Sharpe and Sortino ratios depend on the chosen risk-free rate. When interest rates change significantly (as they did between 2020 and 2023), the same portfolio produces different Sharpe ratios even if its actual returns and volatility are unchanged.
Manipulation vulnerability: Strategies that sell insurance-like instruments (such as far out-of-the-money put options) can produce artificially high Sharpe ratios during normal market conditions. These strategies collect small, steady premiums and appear low-risk until a tail event triggers a catastrophic loss. The Sharpe ratio, measured during the calm period, dramatically overstates the risk-adjusted performance.

Practical Considerations

Calculating and Interpreting Metrics

Several practical details affect how risk-adjusted metrics should be calculated and interpreted:

Annualization: Returns and risk measures should be annualized for comparison. Monthly Sharpe ratios are multiplied by the square root of 12 to annualize. This scaling assumes that monthly returns are independent, which is approximately but not exactly true. For strategies with significant autocorrelation (where returns in one month predict returns in the next), annualized Sharpe ratios can be misleading.
Sufficient data: Risk-adjusted metrics require enough data to produce statistically meaningful estimates. A Sharpe ratio calculated from 12 months of returns has wide confidence intervals. Three to five years of monthly data is a common minimum for meaningful Sharpe and Sortino ratios. The Calmar ratio inherently requires at least three years because it uses the maximum drawdown over that window.
Fees and costs: Risk-adjusted metrics must be calculated net of all fees and transaction costs. A gross-of-fee Sharpe ratio overstates the investor's actual experience. When comparing strategies with different fee structures, the net-of-fee comparison is the only fair one.
Leverage effects: Leverage amplifies both returns and volatility, leaving the Sharpe ratio approximately unchanged. This means a levered and unlevered version of the same strategy can have similar Sharpe ratios despite very different risk levels. The Calmar ratio is more informative in this case because leverage disproportionately increases maximum drawdowns.

Common Benchmarks for Context

Risk-adjusted metrics are most useful when placed in context. Without a reference point, knowing that a strategy has a Sharpe ratio of 0.8 provides limited insight. Typical long-term benchmarks (these vary significantly by period):

U.S. large-cap equities (S&P 500): Historical Sharpe ratio of approximately 0.4 to 0.5 over long periods. Individual years vary widely, from strongly negative to well above 1.0.
U.S. aggregate bonds: Historical Sharpe ratio of approximately 0.3 to 0.4, with lower volatility but also lower excess returns than equities.
Balanced portfolios (60/40): Typically produce Sharpe ratios in the range of 0.4 to 0.6, benefiting from diversification between equities and bonds.
Active hedge fund strategies: Top-quartile managers have historically targeted Sharpe ratios above 1.0, though realized ratios vary substantially and survivorship bias inflates reported averages.