Levent Guntay
Asst. Prof.
Faculty of Business
PREDICTION ON IMPLIED VOLATILITY SURFACES
This study aims to investigate the predictive power of time series and machine learning models, including ARIMA, PCA-VAR, PCR, and Feedforward Neural Networks (FNN), in modeling the implied volatility surfaces of five emerging market currencies (TRY, INR, MXN, ZAR, and BRL) against the USD.
The research evaluates model performance based on the Root Mean Square Error (RMSE) metric using both the Expanding and Rolling Window methods.
The findings reveal that the machine learning models FNN and PCR outperform traditional models such as ARIMA and PCA-VAR, especially for currencies that exhibit lower volatility.
The study emphasizes the importance of model selection in financial forecasting and suggests that the inclusion of macroeconomic or geopolitical factors can further improve the precision of the forecast
Project Year: 2025
Project Partner: This is academic research project joint with Emrah Ahi, Eren Akansel, Levent Güntay, based on the MS Thesis of Eren Akansel
