Logo
Munich Personal RePEc Archive

Fiscal sustainability among Brazilian states: a Markov-switching, unsupervised learning approach

Castro, Hedmus and Alexandre, Michel and Costa Filho, João Ricardo (2026): Fiscal sustainability among Brazilian states: a Markov-switching, unsupervised learning approach.

[thumbnail of MPRA_paper_129276.pdf]
Preview
PDF
MPRA_paper_129276.pdf

Download (466kB) | Preview

Abstract

The purpose of this paper is to identify patterns of fiscal sustainability among Brazilian states, considering regime changes. To this end, we employ two distinct approaches: the Markov-switching autoregressive (MSAR) model and k-means clustering. Our dataset comprises quarterly data on the primary balance of Brazilian states. Our results show that Brazilian states transition between two fiscal regimes characterized by positive or negative fiscal results. Furthermore, Brazilian states can be grouped based on their shared characteristics. Finally, we demonstrate important correlations between fiscal sustainability variables, such as performance in a given fiscal regime and transition probabilities. These results will contribute to a more accurate assessment of the fiscal dynamics of Brazilian states.

Atom RSS 1.0 RSS 2.0

Contact us: mpra@ub.uni-muenchen.de

This repository has been built using EPrints software.

MPRA is a RePEc service hosted by Logo of the University Library LMU Munich.