Publication Type

Master Thesis

Publication Date



Following Phillips, Wu and Yu (2007), this paper extends their bubble detecting work to several Greater China stock markets. Two alternative bubble detecting methods, the forward recursive ADF tests raised by Phillips et al. (2007) and the modified version, forward rolling ADF tests, are implemented and compared. Monte Carlo simulations are performed to determine the critical values of the ADF statistic under different sample size. Empirical results demonstrate that only rolling ADF tests are successful in detecting rational bubbles by overcoming the problem of periodically collapsing bubble. As we have expected, bubbles in China Mainland stock market are detected. Out of our expectation, significant and long standing bubbles are also found in Hong Kong, Taiwan and Singapore stock markets. However, the styles of rational bubbles in different stock markets are different. Differences between the transition stage of China Mainland and the mature stage of other Greater China economies should be one important reason that leads to the different stock market speculative behaviors during the same period. At last, the potential time when bubble begins to collapse is investigated.


bubble collapse, bubble size, forward recursive ADF tests, forward rolling ADF tests, Rational bubbles, speculative behavior

Degree Awarded

MSc in Economics


Asian Studies | Finance | Portfolio and Security Analysis


TSE, Yiu Kuen

Creative Commons License

Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.