This research is motivated by the conditions of uncertainty that characterize today's economic and financial environment in emerging markets, particularly in Argentina, Brazil, Chile and Mexico: Argentina, Brazil, Chile and Mexico. An operational, practical and functional instrument of "early warnings" (of recession or expansion) is developed to estimate the probability of instability and local financial vulnerability. Use is made of third generation non-linear and multiple equilibrium forecasting models. The modeling strategy includes a unit root analysis and cointegration of fundamental macroeconomic variables, which allows for a better level of reliability in the variables to be included, that is, to identify those variables that are more effective for reasons of causality than chance, avoiding the inclusion of only those with spurious r2. The discrete choice probit, logit, dynamic, multiple equilibrium models are also used. This text is of interest to those readers interestedin learning more about this interesting and current topic, which is a vast and promising subject of analysis.