Verification of Causality through VAR and Intervention Analysis: Econometric Modeling on Budget Deficit and Trade Deficit in Nepal

Abstract

Direction of causality between budget deficit and trade deficit, which is popularly known as Twin Deficit Hypothesis (TDH), has been tested in this paper covering the period 1964-2004. Stationarity, co-integration, and error correction tests have been performed as fundamental groundwork on real-term datasets. Datasets are found to be stationary at first difference. Long-run relationship (cointegration) among model variables is found at first difference. Long-run stability has been supported since short-run dynamics indicated converging pattern. Residual tests and conventional Granger Causality tests suggested trade deficit has been Granger Caused by the budget deficit. This initial gesticulation has further been reinforced by the vector autoregressive (VAR) modeling and intervention analysis (impulse response function and variance decomposition) also as it has reconfirmed unidirectional causality from budget deficit to trade deficit indicating need of a policy revisit regarding efficient public expenditure management, export-led growth and strategic capital formation with the help of revised fiscal, monetary and financial policies in the present globalization context.