Managing portfolio managers : the impacts of market concentration, cross-sectional return dispersion and restrictions on short sales
Thesis (PhD)--Stellenbosch University, 2012.
The impacts on the active management of investment portfolios of a) market concentration, b) cross-sectional return dispersion and c) restrictions on short sales are explored in this thesis. The context is the fund sponsor’s management of their investment managers in a South African equity investment environment. Some of the findings here are developed analytically and some make use of multiple simulated investment views and their corresponding optimal portfolio solutions to document the size and nature of the inefficiencies that are created by these three factors. The cross-sectional volatility of asset returns in an investment universe represents a carrying capacity for active risk taking: the higher the cross-sectional volatility, the greater the opportunity for active risk taking. Cross-sectional volatility is shown to be an important consideration when setting active risk targets. It is shown that, to remain efficient, active risk should be reduced during periods of low cross-sectional dispersion and vice versa. The sensitivity of active risk estimates to changes in the cross-sectional dispersion of their investment universe is demonstrated and sponsors should therefore exercise caution when reacting to changes in the active risk estimates of their funds. Cross-sectional volatility is shown to be time-varying and is related to similarly varying dispersion in realised fund returns. The ex post performance of competing portfolio managers therefore require correction for this heteroscedasticity and an effective weighted adjustment is recommended. Active managers can only fully express their views in an environment where their mandated conditions accommodate their conviction and level of risk taking. The short sale restriction is shown to be materially binding when applied to a concentrated benchmark such as the ALSI where only a few of the stocks comprise most of the total investment weight. The more concentrated the benchmark and the higher the active risk target, the wider the distribution of individual asset weights in the portfolio will be and the more binding the weighting constraints will be. It is shown that constraints on short positions are more binding on assets with small weightings in the benchmark illustrating the asymmetrical sub-optimal effect of these constraints when they are applied uniformly across the investment opportunity set. It is argued that requiring long-only managers to increase their active positions and/or active risk in a concentrated investment environment further constrains them in their ability to express their best investment view and increases their competitive disadvantage relative to unconstrained funds taking similar risk. The research presented in this thesis measures the nature and size of the impacts of the market concentration, cross sectional return dispersion and restrictions on short sales that are implied by the investment mandate on the quality of the investment portfolio, providing analysis and techniques which can inform and improve the quality of the relationship between fund sponsor and fund manager. The more appropriate the investment mandate and the monitoring of the fund’s performance subject to this mandate, the more effective the manager’s risk-taking on behalf of their investors will be. This is the principle that this research aims to serve.