Rebalance Timing Luck: The Difference Between Hired and Fired

Posted: 31 Jan 2019 Last revised: 7 May 2019

See all articles by Corey Hoffstein

Corey Hoffstein

Newfound Research

Justin Sibears

Newfound Research

Nathan Faber

Newfound Research

Date Written: October 1, 2018

Abstract

Prior research demonstrates that performance can be highly sensitive to index construction choices related to rebalance dates. The authors define the measure of “rebalance timing luck” as the standard deviation in returns between identically managed investment portfolios that are rebalanced on different dates (“sub-indexes”). Using a simple fixed-mix strategy, they identify that rebalance timing luck can have significant implications for realized results. Statistical tests imply that ex-ante expected returns for the sub-indexes are identical and realized performance differences are not expected to mean-revert over time. The authors demonstrate that investors can minimize the impact of rebalance timing luck through equal-weight exposure to the sub-indexes. They propose an ex-ante model to estimate the magnitude of rebalance timing luck and find that the process equally-weighting across N sub-indexes reduces rebalance timing luck by 1/N.

Keywords: Rebalancing, Calendar-Based Rebalancing, Partial Rebalancing, Portfolio Construction, Fixed-Mix Strategies, Staggered Rebalancing

JEL Classification: G11

Suggested Citation

Hoffstein, Corey and Sibears, Justin and Faber, Nathan, Rebalance Timing Luck: The Difference Between Hired and Fired (October 1, 2018). Available at SSRN: https://ssrn.com/abstract=3319045 or http://dx.doi.org/10.2139/ssrn.3319045

Corey Hoffstein (Contact Author)

Newfound Research ( email )

380 Washington Street
2nd Floor
Wellesley, MA 02481
United States

HOME PAGE: http://www.thinknewfound.com

Justin Sibears

Newfound Research ( email )

425 Boylston Street
3rd Floor
Boston, MA 02116
United States

Nathan Faber

Newfound Research ( email )

425 Boylston Street
3rd Floor
Boston, MA 02116
United States

Do you have negative results from your research you’d like to share?

Paper statistics

Abstract Views
5,417
PlumX Metrics