Modeling The Volatility Of EUR/CHF Exchange Rate in The Albanian Market: An Approach Using Latent Likelihood

Abstract

Objective: The objective of this study is to estimate the volatility of the EUR/CHF exchange rate in Albanian market, with the purpose of forecasting it within a short time frame of one day. This is particularly important given the interesting fluctuations observed in the past year. Such information is crucial for economic agents who face exposure to this specific exchange risk, as both EURO and CHF are widely used and strong European currencies within the country. Theoretical Framework: This topic covers studies that incorporate modifications to standard deviation errors in autoregressive models and those that address autocorrelated errors using varying variance models. Method: The study employs the ARMA (1;1), ARCH (1;1), and GARCH (1;1) models to calculate the EUR/CHF volatility with a 95% confidence level. It employs daily EUR/CHF exchange rates, denominated in ALL, from May 4, 2022, to April 28, 2023. This period encompasses a total of 245 observations. Results and Discussion: The results suggest that both the ARMA (1;1) and ARCH (1;1) models were inadequate in accurately estimating the volatility of the EUR/CHF exchange rate. This discrepancy is likely due to latent factors that impact volatility within the Albanian market. In contrast, the GARCH (1;1) model proved effective in capturing EUR/CHF exchange rate volatility. Research Implications: The GARCH (1;1) model is implemented using the moving window method over a one-day timeframe. To forecast the EUR/CHF exchange rate on 02/05/2023, we utilize 244 data points spanning from 04/05/2022 to 28/04/2023. This process is repeated for the entire period from 02/05/2023 to 30/04/2024. Originality/Value: The study evaluates the statistical robustness of three dynamic models to accurately estimate EUR/CHF exchange rate volatility in the Albanian market, filling a gap in existing literature. It provides future predictions for the EUR/CHF exchange rate within a one-day timeframe by utilizing findings from the precise GARCH (1;1) model. Moreover, the research conducts to Value at Risk (VaR) calculations, enhancing its relevance for informing risk management strategies in the Albanian market.