Analysis of bull and bear phases of the Singapore stock market

• Main Authors: Wang, Tiantian, Yang, Na
• Other Authors: Low Chan Kee
• Format: Final Year Project
• Language: English
• Published: 2010
• Subjects: DRNTU::Business::Finance::Stock exchanges
• Online Access: http://hdl.handle.net/10356/38549
• Institution: Nanyang Technological University

This project studies the alternation of bull and bear markets of Singapore stock market. Using historical daily Straits Time Index (STI) from 25/1/2000 to 24/8/2009 and weekly STI from 28/12/1987 to 24/8/2009, this project investigates two approaches to simulate the bull and bear runs, which are compared with the actual bull and bear runs. The first approach replicates the distributions of bull and bear durations by employing three parametric models, namely, the random walk, GARCH and EGARCH. Random walk model is rejected since it cannot capture the asymmetric features of the observed bull and bear durations. Although GARCH model is better than random walk model, large number of bull and bear durations suggests that the status alternate too frequently. The modified EGARCH model eliminates the above two drawbacks, but it still fails to capture the duration of the longest bull run observed in the data. The second approach applies the idea of Hidden Markov model, which classifies the markets into two states. Although the model has some potential, the results are not consistent with the actual data.