Exchange Rate Modeling and Forecasting (USD/LYD) Using Box- Jenkins (ARIMA) Models for the Period 1980-2024

Abstract

The exchange rate of the Libyan dinar against the US dollar has witnessed significant fluctuations and a substantial increase in recent years due to economic, political, and security factors. This has had a direct impact on the macroeconomy and the prices of goods and services. This study aimed to model and predict the exchange rate (US dollar vs. Libyan dinar) from 1980 to 2024 using ARIMA (Box-Jenkins) time series models. The main problem addressed is the scarcity of precise quantitative studies that help understand and forecast the performance of the exchange rate in Libya, especially amid economic and political volatility and Analysis Data by EViews.

The study relied on historical exchange rate data and followed the ARIMA methodology, which includes the stages of identification, estimation, diagnostic checking, and forecasting. Key results revealed that the time series was non-stationary at its original level but became stationary after taking the first difference. Based on the AIC, SC, and SSE criteria, the optimal model selected was ARIMA(2,1,1).

The study also uncovered a long-term upward trend in the exchange rate, reflecting inflationary pressures and a weakness in the value of the Libyan dinar. It further indicated the presence of heteroscedasticity (ARCH effect), necessitating the use of more advanced models such as GARCH in future studies. The exchange rate was forecasted until 2030, with an expected rate of approximately 6.74 LYD per USD.

Key recommendations include: continuous monitoring of the exchange market, addressing the structural causes of the dinar's weakness, employing GARCH models to improve forecast accuracy, and diversifying the economy to reduce dependence on foreign currencies.