| Abstract [eng] |
Stock market is known as always changing, unstable, risky and dependent on variety of factors. Forecasting the value of an investment portfolio can become a challenge due to market changes. As a result, there is no correct model for predicting the share prices. In this work, investment portfolio consists of five units of Microsoft Corporation and five units of Barclays Bank PLC shares. The value of the investment portfolio and future stock prices are forecasted for 18 days by performing technical analysis and using historical stock prices. Historical stock returns are approximated by orthogonal series density estimator and the independent Metropolis-Hastings algorithm, one of the most prominent Markov Chain Monte Carlo methods, is used to generate random stock returns. In addition, classical Monte Carlo model is also used for prediction of stock values. Orthogonal series density estimate, independent Metropolis-Hastings algorithm and classical Monte Carlo method is implemented in the open source RStudio software and share values computed are compared to historical prices. |
