Assessment of the Relative Efficiency of Bootstrap Resampling Methods for Forecast Uncertainty Quantification in Nigerian Macroeconomic Time Series: Evidence from Inflation and Exchange Rate Data (1960-2025)
DOI:
https://doi.org/10.54536/ajase.v5i2.8771Keywords:
Arima, Bootstrap Resampling, Monte Carlo Simulation, Moving Block Bootstrap, Nigeria Sieve Bootstrap, Stationary Bootstrap, Time Series Forecasting, Uncertainty QuantificationAbstract
Quantifying forecast uncertainty is a prerequisite for responsible use of time series models in Nigerian macroeconomic analysis, yet researchers rarely benchmark resampling schemes used to estimate that uncertainty against alternatives on long-run Nigerian data. This study compares three bootstrap procedures for response time series the stationary bootstrap, the moving block bootstrap, and the sieve bootstrap across five ARIMA data-generating processes of increasing complexity: ARIMA(1,0,0), ARIMA(1,1,0), ARIMA(1,1,1), ARIMA(2,1,1), and ARIMA(1,1,3). The simulation design follows a Monte Carlo framework comparable to earlier benchmarking studies. The same three methods are then applied to two real Nigerian macroeconomic series: annual headline inflation (World Bank/FRED, 1960-2025, 66 observations) and the log of the annual NGN/USD exchange rate (Penn World Table/FRED, 1960-2023, 64 observations), both considerably longer than the short annual series used in prior comparative work of this kind. Across all five simulated data-generating processes and both empirical series, the stationary bootstrap yielded the lowest mean squared error (MSE) of one-step-ahead forecasts; the moving block bootstrap was a close second throughout; the sieve bootstrap was markedly less accurate, particularly for integrated I(1) processes. These results diverge from prior work suggesting block-based methods are dominated by ARMA processes with genuine moving-average components, and from work favouring the sieve bootstrap for autoregressive-dominated series. Possible explanations are discussed, including the sensitivity of the sieve reconstruction step to compounding during re-integration of a differenced series. The scope limitations of a reduced-scale Monte Carlo design are noted, and further replication is identified as necessary before firm practical guidance can be issued to Nigerian forecasting practitioners.
Downloads
References
Adewole, A. I. (2024). Modelling long memory volatilities of Nigeria selected macro-economic variables with ARFIMA and ARFIMA-FIGARCH. Cumhuriyet Science Journal, 45(3), 618–628. https://doi.org/10.17776/csj.1467360
Bergmeir, C., Hyndman, R. J., & Benítez, J. M. (2016). Bagging exponential smoothing methods using STL decomposition and Box–Cox transformation. International Journal of Forecasting, 32(2), 303–312. https://doi.org/10.1016/j.ijforecast.2015.07.002
Bühlmann, P. (1997). Sieve bootstrap for time series. Bernoulli, 3(2), 123–148.
Efron, B. (1979). Bootstrap methods: Another look at the jackknife. The Annals of Statistics, 7(1), 1–26. https://doi.org/10.1214/aos/1176344552
Federal Reserve Bank of St. Louis. (n.d.). Exchange rate (market estimated) for Nigeria (XRNCUSNGA618NRUG) [Data set]. FRED. Retrieved October 6, 2026, from https://fred.stlouisfed.org/series/XRNCUSNGA618NRUG
Federal Reserve Bank of St. Louis. (n.d.). Inflation, consumer prices for Nigeria (FPCPITOTLZGNGA) [Data set]. FRED. Retrieved October 6, 2026, from https://fred.stlouisfed.org/series/FPCPITOTLZGNGA
Künsch, H. R. (1989). The jackknife and the bootstrap for general stationary observations. The Annals of Statistics, 17(3), 1217–1241. https://doi.org/10.1214/aos/1176347265
Lahiri, S. N. (1999). Theoretical comparisons of block bootstrap methods. The Annals of Statistics, 27(1), 386–404.
Liu, R. Y., & Singh, K. (1992). Moving blocks jackknife and bootstrap capture weak dependence. In R. LePage & L. Billard (Eds.), Exploring the limits of bootstrap (pp. 225–248). Wiley.
Oni, L. B., & Akinsanya, T. A. (2026). Analysing the effects of inflation, interest rate and exchange rate on Nigerian economy. American Journal of Applied Statistics and Economics, 5(1), 122–130. https://doi.org/10.54536/ajase.v5i1.7070
Pan, L., & Politis, D. N. (2016). Bootstrap prediction intervals for linear, nonlinear and nonparametric autoregressions. Journal of Statistical Planning and Inference, 177, 1–27. https://doi.org/10.1016/j.jspi.2014.10.003
Politis, D. N., & Romano, J. P. (1994). The stationary bootstrap. Journal of the American Statistical Association, 89(428), 1303–1313. https://doi.org/10.1080/01621459.1994.1047687
Udoudo, K. J., Emele Kalu, I., & Oduola, K. (2023). Impact of liquefied natural gas exports on the Nigerian exchange rate: An ARDL cointegration approach, 2000 to 2021. American Journal of Applied Statistics and Economics, 3(1), 1–14. https://doi.org/10.54536/ajase.v3i1.2216
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Okafor, Joseph Iweanandu, Kerry C.C, Amagoh Maureen N
This work is licensed under a Creative Commons Attribution 4.0 International License.
