Authors: Samer A.M. Al-Rjoub
Addresses: Department of Banking and Finance, Hashemite University, P.O. Box 330195, Zarqa 13133, Jordan
Abstract: This paper adds new methodology to the existing methods used in the literature of early warning systems and signals approach developed to predict crashes. Results show that the probability of extreme outcomes is much higher before the crises than during the crises. Even though the two distributions are far from normal, the distribution tails for returns before crises is much thicker than that of the distribution for returns during the crises. The fact that kurtosis will register high numbers before the crises is a signal of an upcoming crash.
Keywords: financial crises; asset price bubbles; bursting bubbles; early warning systems; US stock market; USA; United States.
International Journal of Trade and Global Markets, 2011 Vol.4 No.1, pp.1 - 12
Available online: 03 Jan 2011 *Full-text access for editors Access for subscribers Purchase this article Comment on this article