Choosing a Window for Rolling Sortino Ratios

Summary

The document raises a methodological choice when calculating a rolling Sortino ratio. One approach rolls the average realized return over a ten-year window but calculates downside deviation across the full history. The other calculates both the return measure and downside deviation over the same rolling window.

It provides no recommendation, formula details, data examples, or empirical comparison, so it does not establish which approach is correct. The distinction is useful because the choices represent different measurement horizons: a full-history downside estimate may change slowly, while a rolling estimate reflects more recent downside behavior. A researcher would need to define the intended interpretation and compare the effects of each choice, including how downside deviation is calculated, before selecting a method. The document is a question rather than a resolved explanation.

Key ideas

• A rolling Sortino ratio can use a rolling return estimate with downside deviation from the full sample.
• An alternative is to calculate both the return and downside deviation over the same rolling window.
• The document does not resolve which windowing choice is preferable.
• The appropriate choice depends on the intended meaning of the rolling risk-adjusted measure.

Tags

# How to create Rolling Sortino Ratio?


# How to create Rolling Sortino Ratio?












There are two methods.

- Only the numerator the average realized return will be rolling (10 years) while the downside deviation is calculated from full time period.

- Both numerator and denominator are rolling.

Which one would be correct?

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.