Time Series Analysis of Stock Returns for Two Pharmaceutical Companies Listed in Chittagong Stock Exchange

Abstract

The primary aim of the study is to analyze and prediction of stock returns for 02 popular pharmaceutical companies namely BEXIMCO and SQUARE pharmaceuticals which are listed in Chittagong Stock Exchange. Generally the effective performance of stock market is one of the major indicators for economic development of a country.  In this study, secondary data on stock index and daily average stock price with a sample period 1st January 2010 to 27th December 2016 for selected 02 popular pharmaceutical companies listed in Chittagong Stock Exchange. Descriptive statistics, important graphs, statistical tests, fitted dynamic time series regression models with ARCH effect are used to complete the analysis. It is found that for both companies , the return occurs high with a high risk and risk is low for the companies with small amount of return. Generally SQUARE pharmaceutical has more gross return than BEXIMCO pharmaceutical. The gross returns for both companies follow the non-stationary but the log returns shows stationarity and the transformed variable log returns is used in the analysis to predict the return for these two companies.  The daily log returns of selected 2 companies confer the normality of the white noise of this variable. It is observed that the average VIF for both companies are less than 10, indicate the not severity of multicollinearity and can use these transform explanatory variables ∆Y_t ,  ∆²Y_t , ∆X_t and ∆²X_t in the model. Significant LM test statistic indicates the situation of having ARCH effect for the log return of both companies. Parkinson’s monthly volatility of both companies also confers the conditional heteroscadisticity in the behavior of the residuals. The dynamic regression model with volatility regression of ARCH(1) and ARCH(2) are employed for the log return of BEXIMCO and SQUARE pharmaceutical respectively. A modified ARDL (2, 2) regression model is proposed for forecasting the log return for BEXIMCO and SQUARE pharmaceuticals. Predicted daily log return for BEXIMCO pharmaceuticals for 28 th December ,2016 is 0.78122,i.e. the gross return is 2.1236 with 1-step ahead volatility is 0.04701, whereas the actual return is 2.087. One can try to analyze the data considering dynamic models such as GARCH, PARCH, ARIMAX, EGARCH model and different dynamic panel data models to predict the data.

Key Words: Stock returns, Parkinson’s Volatility, ARCH model, Modified ARDL model.