Dynamic rebalancing strategies applied to factor portfolios

Abstract

Portfolio rebalancing serves as a critical mechanism for maintaining targeted asset allocations and optimizing risk-return trade-offs. This thesis compares the impact of various calendar- and threshold-based rebalancing strategies in the context of factor investing. Specifically, it scrutinizes the influence of rebalancing frequency, rebalancing threshold, and partial rebalancing proportion on performance. While frequent rebalancing enhances factor exposure and gross returns, it simultaneously increases transaction costs, indicating the presence of an optimal trade-off point. Notably, simple calendar-based strategies demonstrate returns on par with their more intricate, threshold-based counterparts, although the latter exhibit greater consistency in achieving high returns. Consequently, portfolio similarity emerges as a more reliable predictor of returns compared to rebalancing period. Partial yet frequent rebalancing consistently delivers superior performance as stock price changes approximate a Geometric Brownian Motion with a drift component.

Date of Award	26 Oct 2023
Original language	English
Awarding Institution	Universidade Católica Portuguesa
Supervisor	Paulo Alves (Supervisor)