|dc.description.abstract||Style investing is considered the holy grail of investment finance and portfolio management, as it avails numerous options to the portfolio manager to simply and persistently beat the market. This research investigates the impact of style investing on the performance of South African unit trusts, from the view of funds which remain true to their investment styles (consistent funds) against funds which drift their styles (drifters) over the period 2005 to 2014. The study examines the extent to which South African unit trusts drift from, or maintain, the styles stated in their mandates, and also explores which set of unit trusts deliver superior and persistent returns between the style consistent funds and the drifters. The Returns Based Style Analysis (RBSA) model is employed on a sample of 42 South African unit trusts, from seven of the most significant style-based strategies (style indices) on the JSE, to establish the true styles of the funds, that is, whether the funds are correctly classified as stated in their titles. The extent of drift amongst the unit trusts is then ascertained using the Style Drift Score (SDS) method, which derives its existence from the RBSA model. The Style Drift Score is calculated as the square root of the sum of the variances of style weights obtained from the RBSA model. Subsequently, the risk adjusted performances of the funds are evaluated using three models, which are, the Capital Asset Pricing Model (CAPM), Fama-French 3 factor (FF3F) model and also the Sharpe ratio, whereas market timing ability is examined using the Treynor-Mazuy model. Furthermore, performance persistence of the funds is tested using contingency tables over 6 months, one year, two years and three years holding periods and, lastly, the chi-square test is employed to test predictability of future performances based on past performances.
The study finds that the styles of the funds are, on average, correctly classified. With respect to the extent of drift, 62 percent of the funds are found to remain true to their styles (consistent), whereas 38 percent of the funds drift their styles. In evaluating performance, the consistent funds are found to outperform the drifters on two of the three models used (namely, the CAPM and FF3F). However, neither the consistent funds, nor the drifters, are able to successfully time the markets. Additionally, the drifters exhibit a higher relative performance persistence, albeit a negative one (that is, loser-loser persistence), over the short term period (6 months) which diminishes considerably as the holding period lengthens to one year. Persistence disappears completely at two years and three years holding periods. The study does not find any conclusive evidence of the predictability of future returns based on past performances.
It is observed from the results that drift causes an undesirable utility loss to investors as the drifters underperform the consistent funds and also exhibit negative performance persistence. This research finds similar results to the majority of studies done in style investing and concurs with most literature on the South African market. Therefore, the results have implications to both policy makers and the investment industry. Policy makers may have to regulate the unit trust industry more vigilantly to identify drift and also propose regulations to the investment industry for periodic disclosures of funds’ stock holdings in order to curb drift in those funds which claim to be following consistent investment strategies yet stray from their mandates. Likewise, plan sponsors may have to liaise more frequently with portfolio managers to swiftly root out drift and ensure that portfolio managers meet the pre-set investment targets. Regulators of the unit trust industry may also need to advise investors to be more careful, when investing in drifting funds, since drifting is also an investment strategy that is considered profitable under variable economic cycles in the investment industry.||en_US