It is well understood that rebalancing is a necessary step in restoring a portfolio of volatile assets back to its target weights. Whether it is performed periodically or triggered when actual weightings move too far from target, rebalancing a portfolio will naturally lead to selling assets that have outperformed the portfolio and buying assets that have underperformed the portfolio. It is much less widely understood that rebalancing can actually
be a source of return for the portfolio. Despite the fact that this observation dates back to 1982 [Fernholz and Shay] and has been successfully used to manage portfolios for nearly as long, it has come under considerable attack in the recent past by some academics and practitioners. The main arguments used by these detractors are:
- There is no return benefit because the portfolio’s expected wealth does not increase.
- The return benefit exists, but is due to diversification, not rebalancing.
- The return benefit relies on mean-reversion.
These arguments may appear compelling at first glance, but all three are fundamentally flawed. This article explains why.