Detecting Bursa Malaysia’s Long-Term Dependence Of Returns Using Rescaled Range Analysis: (1982-2001)
Loading...
Date
2006-06
Authors
Ooi, Kok Hwa
Journal Title
Journal ISSN
Volume Title
Publisher
Universiti Sains Malaysia
Abstract
This study examines the long memory behavior in Malaysian stock prices by
using a method called Rescaled Range Analysis (RSA). The standard conventional
statistical methods are useful in predicting market movements under the assumption that
the data are independent and identically distributed as well as the relationship between
dependent and independent variables are linear. However, these conventional methods are
only able to identify regular periodic cycles. They fail to detect any cycles when the data
are not independent, the cycles are not periodic and the system is nonlinear dynamic. This
thesis applies Chaos Theory and fractals to explain the financial market phenomenon
where the conventional Efficient Market Hypothesis (EMH) is unable to explain them
effectively. RSA is a nonparametric technique that can distinguish the average cycle
length of irregular cycles in a nonlinear dynamic deterministic chaotic system. We also
introduce a new approach by introducing a moving aspect of RSA. We name this Moving
RSA as MRSA. MRSA uses a moving reference period, a moving starting and ending
period instead of a static reference period. It will examine whether there are changing
long-term memory in stock returns before and after large and abrupt drop (LAD) (or
market crash). In general, we have found long-term memory in stock returns for Bursa
Malaysia. However, the long-term memories are weak in most periods except during the
LAD period where stocks exhibited high volatility in returns. This is against the Random
Walk Theory. We are able to detect some predictable patterns before the market crash.
However, we are unable to forecast the exact date for the next market crash. Hence, Bursa
Malaysia is still weak-form efficient. Lastly, the changes in trading volume do not show
any long-term memories. They have only short-term memories.
Description
Keywords
Bursa Malaysia , Rescaled Range Analysis