Ghana Monthly Macro-Financial Panel Dataset (August 2020–January 2026): Exchange Rate, Remittance Proxy, Inflation, T-Bill Rate, Stock Index, and Oil Price
Description
This dataset contains a monthly macro-financial panel for Ghana spanning August 2020 to January 2026 (66 raw monthly observations), compiled from the Bank of Ghana's Summary of Economic and Financial Data bulletins and cross-checked against the World Bank Pink Sheet for international oil prices. Variables included: - GHS/USD exchange rate (end-of-month interbank rate) - Ghana Stock Exchange Composite Index (GSE-CI, end of month) - 91-day Treasury bill discount rate - Year-on-year headline CPI inflation - Brent crude oil spot price (USD/barrel) - Foreign-currency deposit growth (annual growth rate), used as a proxy for diaspora remittance inflows in the absence of a continuous official monthly remittance series Approximately 65-70 percent of observations are directly transcribed bulletin entries; the remainder, concentrated in three identifiable gap windows, are filled by linear interpolation and flagged in the accompanying *_Real_Obs indicator columns. The dataset underlies the empirical analysis in "Compensatory, Not Countercyclical: A Quantile and Wavelet Analysis of Diaspora Remittances and Ghana's Macro-Financial System" (Amankwaah, University of Cape Coast), which applies quantile regression, quantile-on-quantile regression, and wavelet coherence/multiple-correlation methods to test whether remittances behave as a countercyclical exchange-rate stabilizer or a compensatory, consumption-smoothing flow. After the log-return and first-difference transformations described in the paper (removing the first two months), the estimation sample is N = 64.
Files
Steps to reproduce
1. Source bulletins: five Bank of Ghana Summary of Economic and Financial Data bulletins, spaced across August 2020–January 2026, were used to extract end-of-month values for the exchange rate, GSE-CI, T-bill rate, CPI inflation, and foreign-currency deposits. 2. Brent crude oil prices were cross-checked against the World Bank Pink Sheet for the same months. 3. Missing months within three identified gap windows were filled by linear interpolation between adjacent observed values; these are flagged in the *_Real_Obs columns (1 = directly transcribed, 0 = interpolated). 4. For estimation, the exchange rate, GSE-CI, and Brent crude series are converted to monthly log returns; the T-bill rate and CPI inflation are converted to first differences; the foreign-currency deposit growth series is used both in levels (for wavelet analysis) and as a first difference (for quantile regression and OLS). These transformations drop the first two months, yielding N = 64 for the estimation sample. 5. Full transformation formulas and equation references are documented in Table A1 (Appendix A) of the accompanying manuscript.
Institutions
- University of Cape CoastCentral, Cape Coast