The Mythology of Rebalancing: A Random Walk down Performance and Risk Management
33 Pages Posted: 17 Jan 2017
Date Written: October 1, 2016
dvisors and target date funds. One of the appealing aspects of these platforms and products is that they delegate all investment decisions to the asset manager (who is believed to be more sophisticated) and away from the asset owner (who is typically unsophisticated). The appeal of these products and platforms is driven by their ability to make effective forecasts of assets, derive an effective asset allocation, and then to rebalance the portfolios on a periodic basis to this target asset allocation. In this paper, we will just focus on the robo-advisors and argue that one key activity underlying these platforms and even practiced very broadly in institutional investing and defined contribution funds, naïve rebalancing, is predicated on bad theory and myths and fails the key test of whether it truly improves performance and/or risk management. We will demonstrate that many previous studies of rebalancing examine this issue from a simplistic perspective of a two-asset portfolio and a simple measure of risk, volatility. When we examine these strategies from the perspective of a multi-asset portfolio and a broader set of risk statistics (that a sophisticated investor would apply), then these naïve rebalancing strategies are nothing more than a form of poor market timing and the performance is a coin-toss and the risk profile of the portfolio is typically worsened. Once these flaws of naïve rebalancing are exposed, we suggest that investors would be well advised to adopt a more intelligent form of rebalancing, where an intelligent analysis is conducted of the relative attractiveness of assets to then set allocations within a client’s policy ranges. This approach has a higher likelihood of improving performance and risk management and the ideas to implement such a program are in the public domain.
Keywords: Range Rebalancing, Calendar-Based Rebalancing, Intelligent Rebalancing, Robo-Advisors
JEL Classification: G11, G12
Suggested Citation: Suggested Citation