Purpose: The research is related to solvency forecasting methods and their models, their possible application and determination of precision. The objective of the study is to explore available models, examine whether those that are including cash flow indicators are more precise in solvency prediction than those who miss them. The aim of study is to identify the most relevant, robust and accurate warning mechanism of corporate failure for the manufacturing companies' stakeholders to forecast potential problems. Design/methodology/approach: The study is based on data from randomly selected 131 Latvian manufacturing companies with an annual turnover of 1 million EUR to 3 million EUR over the period of 2011 to 2016. The model validation is performed by estimating two types of errors and using receiver operating characteristic (ROC) curves. Findings: Solvency prediction models provide uneven results; h however, there are solvency models that are successfully adaptable for the purpose of insolvency prediction. Research limitations: Not all companies have to prepare cash flow statements; therefore part of the data of the sampled companies was not usable. Although financial data of all Latvian manufacturing companies that became insolvent in 2016 were used, the number of insolvent and solvent companies is asymmetric. Practical implications: The results of the study comprise the authors̕ considerations of the models as recommended, possible and not recommended for predicting insolvency of Latvian manufacturing enterprises.