Markov Chain K-Means Cluster Models and Their Use for Companies’ Credit Quality and Default Probability Estimation

This research aims to determine the existence of inflection points when companies’ credit risk goes from being minimal (Hedge) high (Ponzi). We propose an analysis methodology that determines probability hedge credits migrate speculative and then Ponzi, through simulations with homogeneous Markov chains k-means clustering method thresholds migration among clusters. To prove this, we used quarterly financial data a sample 35 public enterprises over period between 1 July 2006 28 March 2020 (companies listed on USA, Mexico, Brazil, Chile stock markets). For simplicity, make assumption no revolving for companies they face their next payment only operating cash flow. found Ponzi (1) have 0.79 average default, while ones had (0) 0.28, (−1) 0.009, which are inflections point were looking for. Our work’s main limitation lies in not considering entities’ behavior granting altered states (credit relaxation due supply excess).