Inflation is a delicate macroeconomic agent to deal with since it directly affects citizens' lives. In the case of the index of variation in food prices, such an effect is much more sensitive, both for companies in the food sector and for the population in general. Thus, it is of interest to be able, with some degree of certainty, to obtain relevant information on the index of variation in food prices. In turn, the SIPSA (DANE unit) is the government agency in charge of dictating the majority supply centers in the country; the prices with which the products must be traded. In the present work, the problem of predicting the in fl ation of food from a SIPSA index proposed by the author is addressed. Two methodologies are addressed for this purpose. Namely: integrated self-regressive models of seasonal moving averages with exogenous variables (SARIMAX), self-regressive structural vector models (SVAR) and error correction vector models (VECM).