Overlapping and Non-Overlapping Returns in Variance-Ratio Tests

Summary

The document explains the distinction between overlapping and non-overlapping return observations, in the context of preparing a variance-ratio test. Non-overlapping monthly returns use successive, separate periods, such as one return for each calendar month. Overlapping 30-day returns instead begin on successive dates, so adjacent observations share much of the same price history.

The example illustrates the key statistical trade-off: overlapping windows produce more observations, but those returns are dependent because they share days. Non-overlapping returns provide fewer observations and are described as statistically independent in the answer. The practical guidance is to use statistical procedures that account for the dependence when analyzing overlapping returns. The document gives a conceptual explanation rather than a full variance-ratio test specification; it does not describe the required adjustment, assumptions, or how to compare test results across the two sampling choices.

Key ideas

• Non-overlapping returns measure separate, successive intervals such as calendar months.
• Overlapping returns use rolling windows that share observations with neighboring returns.
• Rolling windows provide more return observations than non-overlapping intervals.
• Dependence between overlapping returns must be accounted for in statistical procedures.

Tags

# What is the difference between overlapping and non overlapping returns


# What is the difference between overlapping and non overlapping returns












My prof asked me to make an Variance-Ratio test for overlapping as well as for non-overlapping returns.

What is the difference between overlapping and non-overlapping?

## Answer by Alex C (score 5, accepted)

https://quant.stackexchange.com/a/31086

An example of non-overlapping one month returns: the return in January, the return in February, the return in March, etc.

An example of overlapping 30 day returns: the return from January 1 to January 30, the return from January 2 to January 31, the return from January 3 to February 1, the return from January 4 to February 2, and so on.

There are far fewer non-overlapping returns than overlapping returns. The non-overlapping returns are statistically independent of each other, the overlapping are not. If you are going to use overlapping returns you must use specific statistical procedures that are designed to take the dependencies into account.

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.