The financial markets see a fair amount of usage of predictive technology and automated computer programs that use mining. Data mining utilizes the principle that historic data is a reliable basis for estimating performance in the future. The technology is designed to help investors make better investment decisions by extracting hidden patterns from the available historic data. From the obtained result, we can build various models (using neural networks) to value the company. A company’s valuation can help other companies to weigh up possible mergers and acquisitions. Stock market analysis is widely regarded as a challenging problem in financial time series prediction. This paper discusses the various techniques used for modelling the vast financial data available electronically and the portfolio management methodologies required to help investors and institutions make better investment decisions. Also, it investigates various geopolitical events that place a challenge on successful stock market prediction.