Dynamic Rebalancing

Dynamic Rebalancing is an investment strategy developed in 2008 and has been tracked since January 2009 that involves making tactical adjustments to a portfolio’s asset allocation in response to changes in market conditions. Unlike static rebalancing approaches that follow a predetermined schedule, Dynamic Rebalancing aims to optimize returns by capitalizing on short-term opportunities. Research shows that while strategic asset allocation is the main driver of long-term returns, routine rebalancing to maintain allocation targets has the next biggest impact. Dynamic Rebalancing provides guidance on when to rebalance (using market momentum signals) and by how much (based on the strength of asset moves). Implementation of this strategy began in 2012 for managing over $1 billion in Custom Target Date for 401k plan clients at FiduciaryVest where I was able to add incremental value over quarterly rebalancing.

The strategy tracks 14 asset classes on a monthly basis across 5 time frames (3, 6, 9, 12, and 24 months). Distribution graphs help you visualize the current market conditions relative to each asset class’s return history to identify outlier signals warranting rebalance action. More recently, data on spreads between asset classes (e.g. large cap vs small cap stocks) allows tactical overweighting/underweighting. Applying judgment helps to avoid over-trading on “whipsaw” signals under extreme market conditions.

Simulated results based on a 60/40 stock/bond portfolio show Dynamic Rebalancing would have added 14.1% in cumulative value since 2009 compared to quarterly rebalancing. Value was added more often than subtracted, with smaller draw-downs during major market events.

Key aspects include:

  • Uses 6-month timeframe most frequently for strategic signals for rebalancing between stocks and bonds, while using longer intervals for tactical equity decisions
  • Signals when market positioning exceeds historical 5th/95th percentiles
  • Rebalance to target at 5th/95th percentiles, under/overweight with incremental steps beyond 2 standard deviations
  • Requires confirmation signal before acting
  • Applies judgment on divergent signals

There may be some limitations for real portfolios versus simulated results since policy ranges may constrain signal impact, real-world portfolios may already be positioned such that they cannot capitalize on signals, or high volatility could cause successive signals that may not make sense to act on again.

In summary, Dynamic Rebalancing aims to improve upon standard rebalancing approaches by providing a rules-based methodology for responding to market conditions. It delivered both simulated and actual outperformance while moderating drawdowns.