The development of an alternative rebalancing strategy for a pension fund

Thesis (Ph.D. (Risk Management))--North-West University, Potchefstroom Campus, 2007.Portfolio managers typically have two choices: let the portfolio drift with the markets or give direction to the portfolio by rebalancing according to a target allocation. Simply allowing the portfolio to drift with the markets without rebalancing is no longer appropriate. This leaves portfolio managers with the second option of giving direction to the portfolio according to a target allocation. 'This can be achieved by implementing an existing rebalancing strategy. Portfolio managers typically rely on ad hoc rebalancing strategies that are either calendar-based, such as monthly or quarterly rebalancing, or volatility-based, such as rebalancing whenever the asset ratios are more than 5% from the target. In the volatility-based rebalancing strategies, the percentage from the target indicates when rebalancing should occur, and the percentage relies on rule of thumb or the use of historical data. The problem with using historical data is that it will not necessarily predict appropriate ranges for the future. These ranges give the indication of when rebalancing should occur. This indicates a need to develop a well-defined rebalancing strategy that assists the portfolio manager to manage a portfolio. Such a rebalancing strategy should be easy to implement. The aim of this research was to develop and implement a well-defined rebalancing strategy that is adjustable over time to assist portfolio managers to maintain their portfolios in line with the objectives and risk aversions of the trustees' of a pension fund. The study introduced the concept of a portfolio drift strategy to set the scene for the different rebalancing strategies. This was to emphasise the importance for a portfolio manager to adopt a rebalancing strategy for a pension fund. An overview is provided of the five broad rebalancing strategies followed by some advantages and disadvantages of each. Certain rebalancing strategies were used to explain the benefit of rebalancing and the cost of rebalancing. The study investigated different methodologies used to identify when the portfolio manager should rebalance and how far back the portfolio manager should rebalance. The study focused on Masters' range rebalancing strategy and this strategy was used as the benchmark strategy for the study. The study summarises the benchmark strategy that will be used to evaluate the alternative rebalancing strategies. A selective number of performance measurement tools were used to evaluate the benchmark strategy. This criterion was used to compare the alternative rebalancing strategies. The study introduced a new decision-making method that has the same return as the benchmark strategy but the risk of the portfolio is lower. The decision-making method is called the second difference method (SD-method). The SD-method builds on Masters' range rebalancing strategy by eliminating certain implicit assumptions made by Masters. The elimination of Masters' implicit assumptions led to an increase in the complexity of the SD-method of the procedure to rebalance. The complexity was eliminated using a computer program formulating the rules of the SD-method. The study developed a new method called the Cusum Test method (CT-method) with its own assumptions and risk specifications to identify when rebalancing should occur. The new method discussed in the study was less complex to implement and, in contrast to Masters' range rebalancing, the ranges were adjustable over time. The CT-method outperformed Masters' range rebalancing strategy and the SD-method on a risk-adjusted return basis.Doctora