1. UKM Learning and Research Repository
2. Master Thesis / Tesis Master
3. Faculty of Science and Technology / Fakulti Sains dan Teknologi

Title:	Huber's robust estimation method in asset pricing of Shanghai stock exchange
Authors:	Wang Jianlong (P66028)
Supervisor:	Saiful Hafizah Jaaman, Prof. Dr.
Keywords:	Stock exchange Huber's robust estimation Shanghai Dissertations, Academic -- Malaysia
Issue Date:	8-May-2015
Description:	The accuracy and efficiency of pricing stock returns is an essential component of investment and modern economy. Shanghai stock exchange (SSE) established in 1990 creates interest in researchers to determine its efficiency and asset pricing using capital asset pricing model (CAPM). Previous studies indicate that the CAPM estimated beta varies with time. Non-stationary beta problem reduces the usefulness of CAPM as a practical assessment tool to manage portfolio's risk. Contrarily, arbitrage pricing theory (APT) provides multiple risk factors in explaining expected return. Normally, the pricing models are employed based on the ordinary least square (OLS) estimator. OLS estimator for a linear model is very sensitive to extreme values. As a new emerging stock market, SSE is influenced by unexpected events or policies giving existence to outliers. OLS thus is not appropriate. Robust regression method is the most suitable alternative to OLS when outliers exist since it is mathematically related to the OLS method. Huber's robust estimator (HRM) takes the advantage of combining robustness with efficiency under the regression model. This research employs and compares OLS and HRM methods in testing the market efficiency and determining the significant factors of SSE. First objective of this study is to test the efficiency of the Shanghai stock market. Applying monthly data from the year 2000 to 2011, the market efficiency test follows the FamaMacbeth cross-section model. The Fama-Macbeth methodgroups the stocks based on the value of estimated beta to create portfolios. Once the portfolios are achieved, the portfolios' return and risk are calculated as the weighted average of the individual stocks. Testing the beta slope of portfolio risks on expected returns, the SSE market is found to be efficient based on both OLS and HRM. The second objective is to determine the significant factors employing the APT model. The significant factors in explaining excess returns differ between OLS and HRM. OLS estimator obtains three micro-economic factors, four macro-economic factors and the constant, meanwhile, HRM estimator obtains two micro-economic factors and five macro-economic factors. In HRM, the constant is not significant indicating that the HRM is adequate to explain the returns. In addition, less standard error of HRM method makes it preferred. Finally, the outlier impact is verified by comparing the OLS estimation performance with HRM following a mixed return model. The mixed return model identifies outlier and regular components. Outliers exist widely in stock returns since there are only four firms without outlier. The results show that the difference of standard error between the two methods is becoming obvious with the increase of the number of outliers. The outliers substantially bias the OLS alpha and standard error. HRM is superior to OLS in reducing the outlier impacts. This research is important as it highlights the significant impacts of outliers in multiple risk asset pricing model of SSE. Market efficiency enables the application of CAPM in China stock market. This research supports the ability of using CAPM in SSE. In this research, seven factors under the HRM estimator considered to be influential to SSE are money supply and quasi money supply, CPI, actual used foreign direct investment, Shanghai mutual fund index, national fiscal expenditure, book-to-market ratio, and cash flow. These significant factors help investors and researchers to choose considerable risks in either practical investing or empirical study. In conclusion, HRM estimator should be used in asset pricing and as investment tool in SSE.,Masters
Pages:	160
Call Number:	QA402.5 .W344 2015
Publisher:	UKM, Bangi
Appears in Collections:	Faculty of Science and Technology / Fakulti Sains dan Teknologi

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.