The production of coffee in Colombia is part of one of the most important lines of its economy, thousands of people depend on this product directly and indirectly who have been affected by the volatility of its price in local and international markets. This study consists of analyzing a predictive model of internal coffee prices that will help mitigate the economic impact of its listing on the stock market. For this, the analysis of time series was used, which allowed to formulate an autoregressive model of a heteroscedastic moving average whose time window includes the price quotes that go from 2010 to 2019, time that included their respective training period and validation. The result is a model that satisfactorily complied with the diagnosis and validation stage. It is concluded from the results that the analysis and construction of the model made it possible to make a prediction of coffee prices for the month of January 2020, the behavior of which presented stability with a slow but sustained decrease in volatility.