Folio Lab/Documentation

Sterling Ratio

A drawdown-based performance measure that improves upon the Calmar Ratio by using the average annual maximum drawdown rather than a single worst-case event, providing a more stable and representative assessment of drawdown risk.

Overview

The Sterling Ratio, attributed to Deane Sterling Jones (1981), addresses a key weakness of measures built on a single maximum drawdown, such as the later Calmar Ratio (Young, 1991): their dependence on one drawdown event. By averaging the annual maximum drawdowns over the evaluation period, the Sterling Ratio provides a more stable and representative measure of drawdown risk.

The original Sterling Ratio included a 10% threshold subtracted from the average annual drawdown in the denominator, acting as a buffer or tolerance level. This threshold was designed to prevent the ratio from becoming excessively large when drawdowns were small. In modern practice, the modified "Sterling-Calmar" variant omits this arbitrary threshold, using only the average annual drawdown in the denominator.

The Sterling Ratio is particularly useful for evaluating managed futures funds, commodity trading advisors (CTAs), and other strategies where drawdown management is a primary concern. By smoothing the drawdown measure across multiple years, it reduces the impact of a single anomalous drawdown event and provides a more stable signal about the strategy's typical risk profile.

Mathematical Formulation

Original Sterling Ratio

The original formulation includes a 10% threshold:

where is the annualized return and is the Average Annual Drawdown. The 10% is subtracted from the average drawdown (note: drawdowns are expressed as negative values, so subtracting 10% increases the absolute magnitude of the denominator, acting as a buffer).

Average Annual Drawdown

The Average Annual Drawdown (AAD) is computed by taking the maximum drawdown within each calendar year and then averaging across years:

where is the number of years in the evaluation period and is the maximum drawdown experienced during year . Each annual maximum drawdown is computed independently using only the price data within that calendar year, with the peak resetting at the start of each year (or alternatively, using a rolling peak from inception).

Sterling-Calmar Variant (Modified)

The simplified modern variant removes the arbitrary 10% buffer:

This variant is more commonly used in practice as it avoids the subjective choice of a threshold value. It can be thought of as a "smoothed Calmar Ratio" that averages drawdown risk across years instead of relying on a single worst case.

What FolioLab Computes

The sterling_ratio field uses the original form with the 10% term. FolioLab holds each annual drawdown as a positive magnitude, so the 10% is added rather than subtracted:

  • The numerator is the CAGR of the full run, the same value the Calmar Ratio uses.
  • Each calendar year of the compound curve gives one maximum drawdown, with the peak reset at the first observation of that year. A move from the last close of one year to the first close of the next does not count toward either year. A partial first or last year counts as a full year in the average.
  • The denominator is never below 0.10, so a portfolio with a small drawdown gets a finite value and not an extreme one.
  • A run with 10 daily observations or fewer gives 0.0.

Relationship to Calmar Ratio

The Sterling Ratio and Calmar Ratio share the same numerator (annualized return) but differ in the denominator. The Calmar Ratio uses the single maximum drawdown over the entire period, while the Sterling Ratio uses the average of annual maximum drawdowns. Because by construction, the Sterling-Calmar variant is typically higher than the Calmar Ratio. The form FolioLab reports adds 10 points to the average, so it can sit below the Calmar Ratio when the maximum drawdown is less than the average annual drawdown plus 10 points. Both forms are more stable over time, as they are less influenced by a single extreme event.

Advantages & Limitations

Advantages

  • More stable than Calmar: Averaging annual drawdowns reduces the impact of a single anomalous event, providing a more representative risk measure.
  • Intuitive risk concept: Like the Calmar Ratio, uses drawdown as the risk measure, which resonates with how investors experience losses.
  • Temporal consistency: Less sensitive to the specific evaluation window than the Calmar Ratio, since no single year dominates.
  • Non-parametric: Makes no distributional assumptions about returns.

Limitations

  • Requires multi-year data: Needs several years of data to compute meaningful annual drawdown averages, making it unsuitable for short track records.
  • Arbitrary threshold: The original 10% buffer is arbitrary and not theoretically motivated.
  • Calendar year dependency: Annual drawdowns depend on how years are defined (calendar year vs. rolling), which can affect results.
  • Drawdown persistence ignored: Does not account for how long drawdown periods lasted or how quickly recovery occurred.
  • Less widely adopted: Not as standardized or widely reported as the Sharpe or Calmar Ratio, reducing comparability.

References

  1. Jones, D. S. (1981). Sterling Ratio methodology. Unpublished industry notes.
  2. Bacon, C. R. (2013). Practical Risk-Adjusted Performance Measurement. John Wiley & Sons.
  3. Lhabitant, F. S. (2004). Hedge Funds: Quantitative Insights. John Wiley & Sons.

Not investment advice. Past performance is not indicative of future results.