Why Equity Momentum Studies Often Skip the Most Recent Month

Summary

The document discusses the practice of excluding the most recent month when measuring stock momentum. It traces the convention to the 1993 Jegadeesh and Titman study, following earlier research by Jegadeesh that identified a one-month return reversal. Skipping the recent month was intended to focus on longer-horizon return continuation rather than the reversal observed at the shortest horizon.

The response says later academic work generally followed that procedure and recommends it as a careful research baseline. It cautions that trying small variations in the skip period may produce apparent improvements that are statistically insignificant or due to random variation. The document does not answer whether skipping a month produces stronger momentum across countries or time periods, and it offers no comparative evidence on one-month versus two-month skips.

Key ideas

• The skip-month convention was used in the 1993 Jegadeesh and Titman momentum study.
• Earlier evidence of one-month return reversal helps explain why recent returns were excluded.
• The procedure aims to measure longer-horizon momentum separately from short-term reversal.
• Small changes to the skip period may appear beneficial by chance rather than reflecting robust gains.

Tags

# Momentum - skipping the most recent month


# Momentum - skipping the most recent month












Many momentum studies skip the most recent month when calculating momentum to account for "reversal effects." On the other hand, I've read online that some people get better results from not skipping the most recent month. And some papers seem to skip two months.

What was the first paper to propose the skip-month? Does the momentum effect generally appear stronger across time and across different countries when skipping a month?

## Answer by nbbo2 (score 6, accepted)

https://quant.stackexchange.com/a/20761

The idea of skipping a month was already in Jegadeesh and Titman 1993. The key academic paper in this area.

Jegadeesh himself (without Titman) discovered a 1-month return REVERSAL effect in 1990, so it makes sense that he would take out 1 month in calculating returns in his later (1993) study. He already knew what happens to stocks that are up a lot over previous month; and wanted to investigate the effect of longer term (3mo, 6mo, 9mo, 12 mo) returns.

Academic researchers coming after have usually reproduced the Jegadeesh Titman procedure. IMHO that is the way careful research should be done. Of course you can also try minor variants "to see if you can do better" but that adds very little value statistically speaking (i.e. non significant, random variations).

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.