Title :
Future trend of the Shanghai stock market
Author :
Wei Wang ; Lin, F.C.
Author_Institution :
Dept of Math & Comput. Sci., Univ. of Maryland Eastern Shore, Princess Anne, MD, USA
Abstract :
In the past decade the dynamic Chinese economy is the fastest growing in the world. If this momentum is maintained, China may replace Japan as the World´s second largest economy in the foreseeable future. In the process, Shanghai metamorphosed into a futuristic, vibrant metropolis with notable cultural and financial institutions of world-class format. It is of interest to monitor its progress. The Shanghai Stock Exchange (SSE) was opened in Dec 1990 as an experiment in China´s transition to a mixed economy. In spite of its brief history, the index is nonetheless a measure of investors´ confidence. We investigate the trend of the Index by inputting the available data into two models: (1) the stochastic SARIMA model and (2) the Backprop Neural Network model. Not surprisingly, these models give divergent predictions. In general, one can characterize the SARIMA model as more optimistic. This is not surprising, since the past performance of the SSE has been bullish. The neural network models, however, give a general downward trend. Comparison with actual data in a 48 weeks forecast period shows close agreement with our predictions.
Keywords :
backpropagation; forecasting theory; neural nets; stochastic processes; stock markets; Backprop Neural Network model; Levenberg-Marquardt algorithm; Shanghai Stock Exchange; Shanghai stock market; cultural institutions; dynamic Chinese economy; financial forecast; financial institutions; future trend; futuristic vibrant metropolis; stochastic SARIMA model; Biological neural networks; Cities and towns; Computer science; Cultural differences; Economic forecasting; Economic indicators; History; Investments; Monitoring; Stock markets;
Conference_Titel :
Neural Information Processing, 2002. ICONIP '02. Proceedings of the 9th International Conference on
Print_ISBN :
981-04-7524-1
DOI :
10.1109/ICONIP.2002.1201908