Forecasting Singapore’s pharmaceutical industry using time series models.

• Main Authors: Lee, Kai Wee., Simandjuntak, Daniel Perdana., Zhuo, Yaohong.
• Other Authors: Choy Keen Meng
• Format: Final Year Project
• Language: English
• Published: 2010
• Subjects: DRNTU::Social sciences::Economic development::Singapore; DRNTU::Social sciences::Economic theory::Macroeconomics
• Online Access: http://hdl.handle.net/10356/35256
• Institution: Nanyang Technological University

In this study on Singapore’s pharmaceutical industry, our objective is to investigate the feasibility of using time series methods to generate one-year forecast for the industrial output based on 204 observations from January 1992 until December 2008. This study is motivated by the lack of published studies on forecasting for the pharmaceutical industry output in Singapore. Time series methods such as the Box-Jenkins and GARCH (General Autoregressive Conditional Heteroscedasticity) methodology were used. In-sample and out-of-sample forecasts were generated using the recursive estimation method and were evaluated based on the root mean square error (RMSE) of their forecasts. The results showed that the ARIMA (2,1,1) with its 1st autoregressive lag removed produced the best forecast using the Diebold-Mariano Statistic with root mean square error as its criterion. However, both of the time series models we have chosen were relatively inadequate in forecasting the pharmaceutical industry output in Singapore. We explained volatility in industrial output and our feedback came mostly from Singapore Economic Development Board and Bristol Meyers Squibb Singapore. Judgmental Forecasting and forecast combinations were suggested as alternative approaches. The report ended with limitations of our study and scope for further research. A regression approach may be feasible, if certain informational requirements can be satisfied.