|
|
 |
| |
|
|
|
Forecasting Pharmaceutical Stock Prices in the Saudi Exchange Using Geometric and Fractional Brownian Motion Models |
|
|
|
PP: 1009-1017 |
|
|
doi:10.18576/amis/200411
|
|
|
|
Author(s) |
|
|
|
Anas Abbas,
Mohammed Alhagyan,
|
|
|
|
Abstract |
|
|
| This study compares four continuous-time stochastic models for forecasting the daily closing prices of two pharmaceutical companies listed on the Saudi Exchange: Saudi Pharmaceutical Industries and Medical Appliances Corporation (SPIMACO) and Jamjoom Pharmaceuticals Factory Company (JAMJOOM). The models under consideration are geometric Brownian motion (GBM), geometric fractional Brownian motion (GFBM), GBM with stochastic volatility (SV-GBM), and GFBM with stochastic volatility (SV-GFBM). We choose these models to examine the efficiency of incorporating memory and stochastic volatility assumptions into GBM. Forecast accuracy is assessed by mean squared error (MSE) and mean absolute percentage error (MAPE). According to the two accuracy instruments, the SV-GFBM specification yields the smallest error values; however, the differences among competing models are modest. This result cannot be conclusive, so the evidence should be interpreted as exploratory and comparative. The contribution of the paper is methodological: it documents how long-memory and stochastic-volatility assumptions affect mean-path forecasts for two Saudi pharmaceutical stocks. This result has ensured the direct positive affection of incorporating the assumptions of long memory and stochastic volatility into the GBM model which agrees with some previous studies. |
|
|
|
|
|
|
|
