A Simulation Study on The Simulated Annealing Algorithm in Estimating The Parameters of Generalized Gamma Distribution

Abstract

The gamma distribution is one of the most widely used statistical distribution in fitting various real life data set in the field of the reliability to fit failure data, because it is sufficiently flexible to deal with decreasing, constant, and increasing failure rates. In this article, the stock investment valuation model has been presented on the assumption of Gamma distribution. The purpose is to analysis profitability of investment in the property development sector in Malaysia based on the investment rate of return on investment. The flow of the investment cash and the accumulation of the share units over the investment period have been presented. it demonstrated the computing of MIRR by the case of a company which experienceed distributing a treasury share dividend, which increases the small portion of investor’s share units instead of receiving the dividend in cash. As the MIRR might exhibit negative values, the Gamma distribution of the transformed MIRR for certain investment periods is developed and the method of moment is applied to estimate the parameters. The study revealed that the transformed MIRR is well fitted with Gamma distribution for four years to eight years investment periods. It is shown that this approach has a better performance return on investment in Malaysia property sector from 2008 to 2018.

Keywords

• simulated annealing algorithm
• generalized gamma distribution
• parameters estimation methods
• log-likelihood function