|
|
 |
| |
|
|
|
GDP Forecasting for Policy: Evaluating ARIMA, Exponential Smoothing, and XGBoost Models for Fiscal, Monetary, and Welfare Planning |
|
|
|
PP: 185-201 |
|
|
doi:10.18576/jsap/150203
|
|
|
|
Author(s) |
|
|
|
Mohammed Wamique Hisam,
Khaliquzzaman Khan,
Mohammad Shahfaraz Khan,
Vanshita Arora,
Amir Ahmad Dar,
Sukriti Kakkar,
Aseel Smerat,
|
|
|
|
Abstract |
|
|
| A key economic metric that represents stability, productivity, and national growth is the gross domestic product (GDP). For academic research as well as for decision-makers at various levels, accurate GDP forecasting is crucial. Three distinct time series models are used in this study to anticipate GDP values for 2022–2026: Exponential smoothing (ES), Extreme Gradient Boosting (XGBoost), and AutoRegressive Integrated Moving Average (ARIMA). The study focuses on the GDP of the United States and India during 2000–2021. Forecasts of GDP are used by governments to create fiscal budgets, investments in infrastructure, and social welfare initiatives central banks like the U.S. and Indian Reserve Banks. Forecasts are used by the Federal Reserve to direct inflation control, interest rate decisions, and monetary policy. Such forecasts are incorporated into loan programs, debt management plans, and worldwide economic monitoring by international organizations such as the World Bank and the International Monetary Fund (IMF). GDP projections are used by companies and investors to guide supply chain planning, trade negotiations, and investment risk management. Three models’ predictive accuracy is assessed in this study, and ES is found to be the most accurate for short-term GDP projections. This study shows that GDP prediction is not only a technical challenge but also a basis for evidence-based governance, economic stability, and inclusive growth policies in both emerging and mature economies by connecting forecasting accuracy to real-world applications. |
|
|
|
|
|
|
|
