Real exchange rate misalignment and exchange rate regime dynamics: Evidence from Sierra Leone
Abstract
This study examines whether Sierra Leone’s succession of exchange-rate regimes since 1970 has been associated with changes in the alignment between the real exchange rate (RER) and its macroeconomic fundamentals. Using an Edwards–Elbadawi framework, the study combines ARDL bounds testing and error-correction modelling, Bai–Perron structural-break analysis, and forecast error variance decomposition to estimate the equilibrium RER and its misalignment over 1970–2024. The model incorporates terms of trade, trade openness, government expenditure, investment, net capital flows, a technology proxy based on real GDP growth, and the interest-rate differential. The results provide evidence of a long-run cointegrating relationship and a strong error-correction mechanism, with approximately 87% of the previous period’s disequilibrium corrected within one year. Although the full-sample average misalignment is close to zero (–0.21%), this conceals substantial variation: the Leone was undervalued in 25 years and overvalued in 30 years, with several annual deviations exceeding 30% in absolute magnitude. Regime-level average misalignment ranges from –26.48% during the liberalised but still managed float of 1986–1989 to +14.71% under the US-dollar peg of 1983–1985, while the 2010–2019 floating-with-managed-features period records the smallest absolute average misalignment (–0.12%). The Bai–Perron analysis further identifies significant breaks in the coefficient linking net capital flows to the RER in 1987, 2000, and 2010, indicating partial parameter instability rather than instability in the entire long-run relationship. Forecast error variance decomposition shows that government expenditure, trade openness, and terms of trade account for increasing shares of RER forecast uncertainty at longer horizons. Overall, the findings indicate that Sierra Leone’s RER misalignment is characterised by recurrent and sizeable deviations across changing policy and macroeconomic environments rather than by a persistent average distortion.