This paper presents a novel conceptual and methodological approach to measuring financial conditions in developing countries drawing on dynamic factor analysis. Our theoretical foundation to construct and interpret the model is based on a Minskian framework of financial instability. Conceptually, instead of analysing financial conditions for individual economies, we cluster various countries with similar financial dynamics into different groups. This has the advantage of alleviating data scarcity and data quality problems. Our methodology also allows for potential regime changes in countries through various specifications for loadings. The paper presents the results of our approach with fixed loadings, which resulted in five different clusters. Most countries fall into a classic boom and bust type of financial cyclicality that renders stable and long-term development financing inherently difficult. We finish the paper by outlining our policy recommendations on the global level as well as targeted measures for each cluster.