The effect of balance of payment components on exchange rate dynamics in Uganda
No Thumbnail Available
Date
2026-05-25
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Uganda Christian University
Abstract
The study assesses the short and long-run impact of the components of the Balance of Payments (BOP) on the exchange rate dynamics in Uganda for quarterly data from 2014Q1 to 2024Q4 (n=44 observations). The study is driven by the fact that Uganda has been facing persistent current account deficits (CA deficit) averaging at 7.9 percent of GDP during the study period, the choice of floating exchange rate regime by Bank of Uganda, the frequent foreign exchange shortage crises, and the imminent transition to commercial oil production. In combination, these factors make a case for using exchange rate stability as the core focus of macroeconomic policy. The study uses an Autoregressive Distributed Lag (ARDL) bounds testing approach to estimate an ARDL(3,3,3,3) error-correction model of the log nominal UGX/USD exchange rate (lnNER) using three regressors retained after conducting the tests for multicollinearity namely, net exports as a percentage of GDP (NX), net Foreign Direct Investments as a percentage of GDP (FDI), and consumer price inflation (INF). The volatility properties are analysed following a GARCH(1,1) specification, while the dynamics between the variables are explored by a VAR(2) specification of the variables, and pairwise Granger causality tests. The bounds test confirms the cointegration (F = 4.819, higher than the upper critical bound of the 5 per cent for the test). The error correction term (ECT) is statistically significant and correctly signed, and means that around 18.6 percent of any short-run imbalance is corrected in each quarter, with the complete adjustment taking place in about 5.4 quarters. Also, in the short run, FDI appreciates by two quarters, which is in line with the portfolio balance theory (coefficient = −8.158, p = 0.003), whereas lagged inflation causes depreciation (coefficient = 0.011, p = 0.024), consistent with the Purchasing Power Parity hypothesis (PPP) that higher inflation erodes the external value of currency. The GARCH(1,1) estimates suggest α + β = 1.043, suggesting that the volatilities are nearly always present due to episodic shocks to external factors, in particular the depreciation in 2015, the COVID-19 disruption and the global surge in commodity prices in 2022. The Granger causality results show that inflation has a significant effect on the causality of net exports (χ² = 7.384, p = 0.025) and has strong bidirectional relationship with NX and INF, and weaker bidirectional relationship with NX and exchange rate, but none of the individual BOP variables Granger-causes lnNER at the conventional significance level, which supports the exchange-rate disconnect literature. The residuals diagnostics show that the model is adequate (Breusch–Godfrey p = 0.189, White p = 0.427, Jarque–Bera p = 0.508, Ljung–Box p = 0.109). The results indicate that the Bank of Uganda needs to adopt a coordinated policy mix of export diversification, inflation targeting and prudent management of foreign reserves to stabilise the shilling in the country's transition to becoming an oil producer.
Description
Undergraduate