PREDICTABILITY OF STOCK MARKET CRASHES: A STATISTICAL APPROACH

The last decades of the last century were marked by a prodigious development of capital markets phenomena modeling. Mathematical modeling, which proved its utility in the study of natural science, was adapted to the economic sphere in order to increase the degree of accuracy of the results and, particularly, predictions. The recent economic and financial crises once again show that the predictions provided by mathematical models were successful especially regarding the past and less regarding future evolutions. This study presents a statistical approach of the predictability of stock market crashes, with applications on the Romanian capital market.