Choosing Rebalancing Frequency for Equity Factors

Summary

The discussion considers how often to calibrate or rebalance equity factors built for research on the S&P. It gives a factor-specific answer: momentum may be calibrated daily, while slower-moving value measures may suit monthly calibration. This implies that one frequency need not fit every factor, even when the goal is general evaluation rather than exact replication.

A second reply highlights the tradeoff in sampling frequency. Daily observations can be noisy, whereas monthly observations may smooth that noise but introduce lag. It consequently suggests daily data as a possible choice, without reconciling that recommendation with the factor-specific guidance. The exchange provides no backtest, factor definitions, calibration procedure, or evidence comparing schedules. Researchers should treat the suggestions as hypotheses and assess them against each factor's signal horizon, noise, and evaluation objective.

Key ideas

• Calibration frequency can depend on the type and time horizon of an equity factor.
• The discussion suggests daily calibration for momentum and monthly calibration for value.
• Daily data may be noisy, while monthly data can introduce lag.
• The replies offer differing general guidance and do not present comparative empirical evidence.
• Rebalancing choices should be evaluated in the context of the factor and research objective.

Tags

# Equity risk factors with daily rebalancing


# Equity risk factors with daily rebalancing












I am building some well known equity factors on the S&P for research purposes. It means those are going to be used for general evaluation purposes but do not need to be replicable. Would it be a good idea to calibrate each factor daily?

## Answer by deb (score 1, accepted)

https://quant.stackexchange.com/a/61273

Depends on what factors you are considering. Things like momentum can be calibrated daily. For Factors like value, it may be a good idea to calibrate monthly.

## Answer by user28909 (score 0)

https://quant.stackexchange.com/a/54381

The issue with daily data is that it is too noisy. The alternative is to use monthly data but there is a lag.

Hence, using daily data might be ur best option.

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.