Model selection for measuring security price performance

Abstract

In this article a bootstrapping routine is used to compare the efficiency of different benchmarks that can be used for measuring security price performance on the Johannesburg Stock Exchange. Four approaches are compared; three benchmarks based on the theory of the Capital Asset Pricing Model, and one, a two-factor benchmark, based on the theory of the Arbitrage Pricing Model. The findings show that, for the JSE, the two-factor approach is superior. This is consistent with prior research into the South African securities market where evidence of two clear factors influencing security returns have been found. The recommendation is that the two-factor benchmark be used for measuring security price performance on the JSE, especially for small samples where only limited benefits can be expected through the central limit theorem.