Inflation forecasting in Thailand: Bayesian Shrinkage and Factor Models in a small open economy

Forecasting inflation in Thailand is challenging because limited time series and strong external exposures create an imbalance between few observations and many potential predictors. We evaluate modern Bayesian shrinkage and factor models, including Horseshoe regressions, factor-augmented autoregressions, factor-augmented VARs, dynamic factor models, and Bayesian additive regression trees, using more than 450 domestic and international indicators. The results show that factor models dominate at