This paper explores the 'Granularity Paradox' in time-series forecasting, where finer temporal disaggregation (e.g., from Monthly to Weekly/Daily) improves in-sample diagnostics and dataset size (N), but degrades out-of-sample accuracy due to recursive error compounding over longer horizons (H). Conversely, coarse aggregation (Annual) eliminates recursive error propagation but reduces data available to estimators.
We formalize this trade-off and benchmark 10 models—spanning naive, statistical, machine learning, and deep learning architectures—across six granularities using a 13-year public procurement dataset. The empirical results reveal a non-monotonic threshold structure: recursive autoregressive and seasonal models degrade substantially under high-frequency forecasting (e.g., Holt-Winters reaches a Test R-squared of -151 and TPFE of 425.85% at the Daily grain), while LSTM traces a U-shaped error curve, worsening from Monthly (19.66%) through Bi-Weekly (35.94%) before overcoming the error propagation penalty at Daily (TPFE of 4.35%, R-squared of 0.66).
Linear Regression remains stable across all granularities (16.3-17.0% TPFE), confirming that the paradox is driven by recursive feedback topology, not model complexity. The results demonstrate that standard pointwise metrics (RMSE, MAE) systematically mask cumulative error propagation, and evaluating forecasts without goal-dependent cumulative metrics produces misleading assessments of model adequacy.
We introduce a consensus-dissensus diagnostic comparing the directional behavior of pointwise metrics against cumulative TPFE across granularities, enabling the identification of models whose standard diagnostics mask systematic error propagation.
Blogger's Review: This paper delves into the impact of granularity choice on model evaluation in time-series forecasting, highlighting potential misleading results when using standard error metrics. By introducing a new diagnostic approach, it better identifies underlying issues in models, offering significant insights for the field of time-series analysis.