Abstract
The tools that are offered to investors in financial markets are fluctuating. As this fluctuation causes losses as well as earnings, it is characterised as a risk for the investor. Especially, fluctuations that may occur in globally important markets and financial instruments have great significance, not just for investor but also for the global economy. Volatility, as a measure of fluctuations taking place in markets, is often used particularly by investors and all economic actors. Therefore, in recent years, future volatility predictions have gained importance. The aim of this research is forecasting future volatility values using the historical data of S&P 500, FTSE 100 and NIKKEI 225 stock market indexes. The progress of historical volatility values in years is presented and generated univariate time series is modelled with artificial neural networks. Future forecasts are done with the obtained model and results are interpreted.
 Keywords: Artificial neural networks, volatility, time series analysis, stock market indexes.
Highlights
Volatility forecasting has gained importance recently and has attracted academics, policy makers, practitioners (Yu, 2002) and every one affected by financial markets
A wide range of literature exists on volatility forecasting
Hu and Tsoukalas (1999) use Artificial neural networks (ANNs) to improve the performance of four conditional volatility models (GARCH, exponential GARCH (EGARCH), IGARCH and MAV) applied to the European Monetary System exchange rates
Summary
Volatility forecasting has gained importance recently and has attracted academics, policy makers, practitioners (Yu, 2002) and every one affected by financial markets. It seems to be possible to model any time series with artificial neural networks. The number of studies modelling and forecasting volatility with ANN is increasing. Hu and Tsoukalas (1999) use ANN to improve the performance of four conditional volatility models (GARCH, EGARCH, IGARCH and MAV) applied to the European Monetary System exchange rates. Donaldson and Kamstra (1997) construct a semi-non-parametric non-linear GARCH model based on ANN to forecast the conditional volatility of stock returns of four international markets. The neural network model uses Euro/dollar and dollar/Yen daily variations, Dow Jones Industrial and Financial Time Stock Exchange daily index returns and daily price variations of oil as inputs. Each time series is modelled with ANNs to forecast weekly volatility values of year 2015.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: Global Journal of Business, Economics and Management: Current Issues
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
