Determination of Arbitrage Pricing Factors in the Malaysian Stock Market
The equilibrium-pricing model using Arbitrage Pricing Theory (APT) has become one of the central models of modern financial theory. However, the APT is too general in determining the factors which influences expected returns. In this regard, the identification of factors is an issue with the most...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English English |
| Published: |
2006
|
| Subjects: | |
| Online Access: | http://psasir.upm.edu.my/id/eprint/262/1/549135_t_gsm_2006_2.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The equilibrium-pricing model using Arbitrage Pricing Theory (APT) has become
one of the central models of modern financial theory. However, the APT is too
general in determining the factors which influences expected returns. In this regard,
the identification of factors is an issue with the most potential to be discussed and
thus it is necessary in the test of the APT. An empirical study on the APT has already
been conducted in developed countries but the identification of factors is still
inconsistent. To add to this problem, little attention has been given in emerging
markets especially in the Malaysian market. This study undertakes to identify the
number of common factors in the asset returns, the number of priced factors in the
expected returns, and the macroeconomic variables that significantly affect the stock
market returns. To identify both factors and the macroeconomic variables, monthly
data was used for a period from January 1991 through December 2001. The analysis
was divided into two sub-periods based on the 1997 Asian financial crisis; namely
sub-period 1 (before financial crisis) and sub-period 2 (after financial crisis). This
will give the research the opportunity to identify the number of common factors and
the number of priced factors in a pre and post economic situation.The objective of this research is to identify the number of common factors, the
number of priced factors and macroeconomic variables that influence the stock
market returns. In this research, Principal Component Analysis (PCA) is used to
identify the groups of macroeconomic variables and to identify the number of
common factors from forty groups of portfolios in the asset returns. The crosssectional
Generalized Least Squares (GLS) regression analysis test is then used to
identify the number of priced factors from the number of common factors identified
earlier using the PCA. The Canonical Correlation Analysis (CCA) is then used to
find the correlation between the Principal Component (PC) scores of the number of
priced factors and the PC scores of the macroeconomic variables. This procedure will
identify which macroeconomic variables significantly affect the stock market returns
before and after the Asian financial crisis.
According to the PCA results, the sub-period 1 suggests that there were six groups of
macroeconomic factors. While after the financial crisis, there were five groups of
macroeconomic factors. The groups of macroeconomic variables in both sub-periods
were factorized from the fourteen Malaysian macroeconomic variables. On the other
hand, the monthly returns of the forty groups were subjected to PCA to identify the
number of common factors. The PCA on stock market returns showed that, the forty
groups consist of ten, eleven, and twelve common factors influencing the asset
returns in sub-period 1. However, in sub-period 2, the PCA on stock market returns
showed that, the forty groups consists of seven, eight, nine, and ten common factors
influencing the asset returns. The cross-sectional GLS regression analysis showed
that there was at least one to three numbers of priced factors in both sub-periods.The CCA suggested that macroeconomic variables and stock market returns in subperiod
1 were significantly correlated at PC scores of the macroeconomic variables;
namely FECON1 (0.4470), FECON2 (0.3214), and FECON3 (0.7791). The results
suggest that FECON1 represents the macroeconomic variables such as the United
States exchange rate, the Singapore exchange rate, and money supply (M2). While,
the FECON2 represents macroeconomic variables such as export, import, industrial
production index, and gross domestic product. FECON3 represents macroeconomic
variables such as oil prices (petroleum) and Japanese exchange rate.
In sub-period 2, the CCA suggested that macroeconomic variables and stock market
returns were significantly correlated at PC scores of the macroeconomic variables;
namely FECON2 (0.6945), and FECON5 (0.4433). The results suggest that FECON2
represents the macroeconomics variables such as the Japanese exchange rate, the
Singapore exchange rate, money supply, trade balance, and oil prices (petroleum).
While the FECON5 represents macroeconomic variables is composed of composite
index. The bottom line conclusion of these empirical findings show that the numbers
of macroeconomic variables are highly significantly influenced by stock market,
namely two PC scores of macroeconomic variables such as FECON3 (0.7791) and
FECON2 (0.6945). They are oil prices (petroleum), and consumer price index before
the Asian financial crisis. While, the Japanese exchange rate, the Singapore exchange
rate, money supply, trade balance and oil prices (petroleum) after the Asian financial crisis |
|---|