BIG DATA ALGORITHMS AND PREDICTION: BINGOS AND RISKY ZONES IN SHARIA STOCK MARKET INDEX

Abstract

Each country with a stock exchange normally calculates various indexes. So is the case
for Malaysia’s Kuala Lumpur Stock exchange (KLSE). FTSE BURSA Malaysia EMAS
Sharia price index (FTBMEMA) is one of its Sharia indexes. In an effort to find which
other indices may forecast this Sharia index, we selected 23 relevant indexes and two
exchange rates. Momentum indicators for short, medium and long term have been
calculated for the variables. The objective of this study is to find predictive indicators
for FTBMEMA out of the population of 188 original and derived variables. Difficulty
arises in reducing the number of variables for regression or other predictive models
like neural networks. In this preliminary study, data mining attribute selection
algorithms along with cross validation criteria have been used, through the use of Java
class library Weka (JCLW), for reducing the number to statistically relevant variables
for our regression estimation in an effort to forecast various performance parameters
for FTBMEMA like performing either in a mean performance range, having jackpots
and bingos or falling into danger zones. Provided the extent of the required predictive
accuracy, the results may bring additional insights for diversifying and hedging
various types of investment portfolios as well as for maximizing returns by portfolio
managers.