Sherif Elbarrad
Financial and Commercial Studies, Vol. 19(Issue 2), 2009, 77-135.
Publication year: 2009

Abstract:

This research was a result of the conclusions and recommendations of the previous two papers regarding the Saudi Stock market. The main goal of this research is to suggest a model to predict the next day’s stock price.

The whole Cement sector is chosen to test the hypothesis on its companies after considering the re-organization which happened to the Saudi Capital Market in 2005 and its possible effect on stock prices. Trading data on the minute level is collected for the period from July 2005 till August 2008.  However, financial data is collected from July 2006 till 2008. The reason is that historical data were needed to calculate some indicators. Data from 2005 till July 2006 were mainly used to prepare for indicators calculations. Big traders’ activities are identified by calculating z-score considering the volume and value of the transactions. If Z-score’s value is greater than 2, then the volume of the transaction which happened during this minute is regarded as a transaction for a big trader. The value and volume of the big transactions during the day are calculated to be used in variables calculations.

More than 40 variables are calculated and used including technical and financial indicators, value at risk, and some variable related to daily stock transactions (stock turnover per day, and stock turnover for big transactions, average transaction volume, average trading price per stock per day…etc.) Such variables are calculated for the whole period for all companies on a daily basis.

The correlation of each variable to the stocks’ prices is tested. The results showed that there is a positive significant correlation between the average stock price for the big transactions and the stock’s price in the preceding day.

A model is built by using the data collected from six companies, and the seventh company’s data is used to test the model.  To build the model, the main variables which had significant correlation with the stock price (based on statistical analysis) are determined. Seven variable are chosen, none of them is related to financial indicators, rather mainly to trading data (such as volume of big transaction in the previous day and value of trading in the previous day) and some technical indicators too. Which indicated that the Saudi Stock market is mainly driven by big traders. Then a step-wise regression is prepared, and a model is developed. The suggested model is then tested on the data of the seventh company in the cement sector. This research showed that the Saudi Capital Market Authority has to keep a closer eye on the activities of big traders.