Conditional forecasting under global commodity-related shocks: Evidence from a Bayesian SVAR for South Africa

Details

Supervisors

Houssa, Romain ; Iacone, Fabrizio

Faculty

Faculté des sciences économiques, sociales, politiques et de communication

Degree label

Master [120] en sciences économiques, orientation générale, à finalité approfondie

Abstract

enThis empirical study examines whether a structurally identified Bayesian SVAR can generate informative shock-conditioned forecast scenarios for South Africa (SA), a commodity-exporting small open economy (SOE). Building on Houssa et al. (2023), the model separates global macroeconomic and commodity dynamics from SA domestic adjustment. It analyses two scenarios: an expansionary world aggregate demand shock and a contractionary world commodity supply shock. Estimated on quarterly data for 1994Q1–2021Q4, with a COVID-volatility correction, the model is evaluated over 2022Q1–2024Q4 against a paired unconditional BVAR benchmark. The results show that shock-conditioned scenarios provide economically interpretable paths and useful information for selected variables, but do not systematically outperform the benchmark. Their predictive content is therefore specific in terms of shock, variable and horizon.