Description:
Early warning systems are too important tools in predicting the crisis in financial institutions say banks and stock markets. A consequence of crashes in a specified stock or stock market is financial crisis. This paper considers designing an early warningsystem based on random walk theory and maximal inequality. First, mathematical tools are presented,and the early warning system is designed, then some real data sets are analysed. The performance of system is evaluated by some different criteria. After it, using a dynamic programming approach, a modified version of mentioned early warning system is proposed. Finally, a conclusion section is given.