Risk Measuring Theory And Empirical Research In The Stock Market

Risk is one of the most hotly discussed topics in finance - so much so that risk measurement and risk management have now become core subjects of financial and investment theory.This paper will both discuss and evaluate risk measurement and provide recommendations for efficiently evaluating risk measurement models. This paper is divided into five parts. Firstly it will introduce the current nature of risk measurement in the domestic and overseas financial sectors and outline the framework of the paper.Secondly it will enumerate existing and varied definitions of risk, distinctly identifying definitions relating to the stock market and will discuss research on risk measurement models.Thirdly it will discuss the backgrounds, assumptions, scope of application and possible improvements for three key risk measurement models; mean variance, capital asset pricing, and the Harlow Model (LPM) and VaR.The fourth part of this paper contains the core content - an analysis of eight stocks performance from 1995 until 2005. The analysis was conducted using the data processing SAS statistical software that deals with the suspension of data, rehabilitation of data, and missing data to effectively improve the accuracy of the model. The SAS algorithms are found to be flexible and manageable. The processing of empirical research works out the CAPM model of an investment portfolio to analyze systemic and unsystematic risk. The objective of this research is to obtain minimal risk of the portfolio by using the MV model. The VaR value will then be calculated according to the weight of the portfolio and further analysis will be discussed covering an historical simulation and Monte Carlo simulation of varying confidence levels. The whole empirical research is systemic and connective.Finally three types of models are comprehensively evaluated according to their risk criterion to ascertain their advantages and disadvantages before offering recommendations for efficiently evaluating risk measurement models.