A Reproducible Forecasting Protocol for African Central Banks Walk-Forward Validation, Diebold–Mariano and Model Confidence Set Tests, Optimal Bates–Granger Combination. An Application to Côte d'Ivoire's CPI (1997-2026)

Boni, Monney Jean (2026): A Reproducible Forecasting Protocol for African Central Banks Walk-Forward Validation, Diebold–Mariano and Model Confidence Set Tests, Optimal Bates–Granger Combination. An Application to Côte d'Ivoire's CPI (1997-2026). Published in: A Reproducible Forecasting Protocol for African Central Banks Walk-Forward Validation, Diebold–Mariano and Model Confidence Set Tests, Optimal Bates–Granger Combination. An Application to Côte d'Ivoire's CPI (1997-2026) : pp. 1-26.

Abstract

Abstract This paper proposes a reproducible inflation forecasting protocol designed for the operational constraints of African central banks, and applies it to the harmonised consumer price index (IHPC) of Côte d'Ivoire over 350 monthly observations spanning January 1997 to February 2026. The protocol rests on four methodological components. The first is walk-forward (rolling origin) validation with 18 forecast origins and four evaluation horizons (h ∈ {1, 3, 6, 12}), which replaces the fixed train-test split still widely used in African applied work. The second is systematic evaluation of forecasts on six metrics (MAE, RMSE, MASE, MAPE, SMAPE, Theil's U), which makes it possible to verify that the conclusions do not depend on the choice of loss function. The third is formal statistical inference, through the Diebold–Mariano test with the Harvey–Leybourne–Newbold small-sample correction and the Model Confidence Set procedure of Hansen et al. (2011), implemented with circular block bootstrap and 5,000 replications. The fourth is the optimal Bates–Granger (1969) combination with covariance-adjusted weights. Five models are compared: ARIMA with automatic AIC order selection, Holt-Winters triple exponential smoothing, random walk with drift, long short-term memory (LSTM) and the neural hierarchical interpolation architecture (NHITS). The application yields three results. First, NHITS achieves a MASE of 0.94 at h = 1, becoming the only model that beats the in-sample random walk benchmark. Second, Holt-Winters is the unique element of the Model Confidence Set at h = 12 and α = 25%, making it the robust choice for annual forecasting. Third, the Bates–Granger combination of NHITS and Holt-Winters produces a measurable 4.84% gain at the quarterly horizon and yields a twelve-month inflation forecast of +1.52%, comfortably inside the BCEAO's 1–3% target band. The complete code is released under the MIT license to facilitate adoption of the protocol by other African economies.

• Item Type: MPRA Paper
• Language: English
• Keywords: inflation forecasting; ARIMA; Holt-Winters; LSTM; NHITS; walk-forward validation; Diebold–Mariano test; Model Confidence Set; Bates–Granger combination; African central banks; Côte d'Ivoire; WAEMU.
• Subjects: C - Mathematical and Quantitative Methods > C1 - Econometric and Statistical Methods and Methodology: General; C - Mathematical and Quantitative Methods > C2 - Single Equation Models ; Single Variables > C22 - Time-Series Models ; Dynamic Quantile Regressions ; Dynamic Treatment Effect Models ; Diffusion Processes; C - Mathematical and Quantitative Methods > C4 - Econometric and Statistical Methods: Special Topics; C - Mathematical and Quantitative Methods > C4 - Econometric and Statistical Methods: Special Topics > C45 - Neural Networks and Related Topics; C - Mathematical and Quantitative Methods > C5 - Econometric Modeling; C - Mathematical and Quantitative Methods > C5 - Econometric Modeling > C53 - Forecasting and Prediction Methods ; Simulation Methods; C - Mathematical and Quantitative Methods > C8 - Data Collection and Data Estimation Methodology ; Computer Programs; E - Macroeconomics and Monetary Economics > E3 - Prices, Business Fluctuations, and Cycles > E31 - Price Level ; Inflation ; Deflation; E - Macroeconomics and Monetary Economics > E3 - Prices, Business Fluctuations, and Cycles > E37 - Forecasting and Simulation: Models and Applications

References

Akiba, T., Sano, S., Yanase, T., Ohta, T., & Koyama, M. (2019). Optuna: A next-generation hyperparameter optimization framework. In Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining (pp. 2623–2631). Alnaa, S. E., & Ahiakpor, F. (2011). ARIMA approach to predicting inflation in Ghana. Journal of Economics and International Finance, 3(5), 328–336. Atkeson, A., & Ohanian, L. E. (2001). Are Phillips curves useful for forecasting inflation? Federal Reserve Bank of Minneapolis Quarterly Review, 25(1), 2–11. Bates, J. M., & Granger, C. W. J. (1969). The combination of forecasts. Operational Research Quarterly, 20(4), 451– 468. Bergstra, J., Bardenet, R., Bengio, Y., & Kégl, B. (2011). Algorithms for hyper-parameter optimization. Advances in Neural Information Processing Systems, 24. Box, G. E. P., & Jenkins, G. M. (1970). Time Series Analysis: Forecasting and Control. San Francisco: Holden-Day. Chen, X., Racine, J., & Swanson, N. R. (2001). Semiparametric ARX neural-network models with an application to forecasting inflation. IEEE Transactions on Neural Networks, 12(4), 674–683. Clemen, R. T. (1989). Combining forecasts: A review and annotated bibliography. International Journal of Forecasting, 5(4), 559–583. Diebold, F. X., & Mariano, R. S. (1995). Comparing predictive accuracy. Journal of Business & Economic Statistics, 13(3), 253–263. Faust, J., & Wright, J. H. (2013). Forecasting inflation. In G. Elliott & A. Timmermann (Eds.), Handbook of Economic Forecasting (Vol. 2A, pp. 2–56). Elsevier. Giacomini, R., & White, H. (2006). Tests of conditional predictive ability. Econometrica, 74(6), 1545–1578. Hansen, P. R., Lunde, A., & Nason, J. M. (2011). The model confidence set. Econometrica, 79(2), 453–497. Harvey, D., Leybourne, S., & Newbold, P. (1997). Testing the equality of prediction mean squared errors. International Journal of Forecasting, 13(2), 281–291. Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory. Neural Computation, 9(8), 1735–1780. Hsiao, C., & Wan, S. K. (2014). Is there an optimal forecast combination? Journal of Econometrics, 178, 294–309. Hyndman, R. J., & Athanasopoulos, G. (2021). Forecasting: Principles and Practice (3rd ed.). OTexts. https://otexts.com/fpp3/ Hyndman, R. J., & Koehler, A. B. (2006). Another look at measures of forecast accuracy. International Journal of Forecasting, 22(4), 679–688. Hyndman, R. J., Koehler, A. B., Ord, J. K., & Snyder, R. D. (2008). Forecasting with Exponential Smoothing: The State Space Approach. Springer. Makridakis, S. (1993). Accuracy measures: Theoretical and practical concerns. International Journal of Forecasting, Page 24 A Reproducible Forecasting Protocol for African Central Banks — Monney (2026) 9(4), 527–529. Makridakis, S., Spiliotis, E., & Assimakopoulos, V. (2018). The M4 Competition: Results, findings, conclusion and way forward. International Journal of Forecasting, 34(4), 802–808. McAdam, P., & McNelis, P. (2005). Forecasting inflation with thick models and neural networks. Economic Modelling, 22(5), 848–867. Nakamura, E. (2006). Inflation forecasting using a neural network. Economics Letters, 86(3), 373–378. Newey, W. K., & West, K. D. (1987). A simple, positive semi-definite, heteroskedasticity and autocorrelation consistent covariance matrix. Econometrica, 55(3), 703–708. Olivares, K. G., Challu, C., Marcjasz, G., Weron, R., & Dubrawski, A. (2022). NeuralForecast: User friendly state- of-the-art neural forecasting models. https://github.com/Nixtla/neuralforecast Oreshkin, B. N., Carpov, D., Chapados, N., & Bengio, Y. (2020). N-BEATS: Neural basis expansion analysis for interpretable time series forecasting. International Conference on Learning Representations (ICLR). Politis, D. N., & Romano, J. P. (1992). A circular block-resampling procedure for stationary data. In R. LePage & L. Billard (Eds.), Exploring the Limits of Bootstrap (pp. 263–270). Wiley. Pufnik, A., & Kunovac, D. (2006). Short-term forecasting of inflation in Croatia with seasonal ARIMA processes. Croatian National Bank Working Paper W-16. Smyl, S. (2020). A hybrid method of exponential smoothing and recurrent neural networks for time series forecasting. International Journal of Forecasting, 36(1), 75–85. Stock, J. H., & Watson, M. W. (2007). Why has U.S. inflation become harder to forecast? Journal of Money, Credit and Banking, 39(s1), 3–33. Suhartono. (2005). New procedures for model selection in feedforward neural networks. Jurnal Ilmu Dasar, 6(2), 104–113. Theil, H. (1966). Applied Economic Forecasting. Amsterdam: North-Holland. Timmermann, A. (2006). Forecast combinations. In G. Elliott, C. W. J. Granger, & A. Timmermann (Eds.), Handbook of Economic Forecasting (Vol. 1, pp. 135–196). Elsevier. Varian, H. R. (1975). A Bayesian approach to real estate assessment. In S. E. Fienberg & A. Zellner (Eds.), Studies in Bayesian Econometrics and Statistics (pp. 195–208). North-Holland. Winters, P. R. (1960). Forecasting sales by exponentially weighted moving averages. Management Science, 6(3), 324–342. Zellner, A. (1986). Bayesian estimation and prediction using asymmetric loss functions. Journal of the American Statistical Association, 81(394), 446–451.