Within the framework of ALM, we understand that immunization is the main method to assure a certain yield in a future date departing from an initial portfolio. As in the crisp case, in a fuzzy environment, one of the goals is the determination of the duration of a bond and of a bond portfolio, when the interest rate is a fuzzy number. After that, we will be able to design an immunization strategy. One of the problems of the immunization strategies derives from the fact that the immunization process allows the variation of the TSIR in a parallel way to the initial, this assumes the hypothesis that the TSIR follows an additive process to the spot rate. Therefore, if this hypothesis is not guaranteed, it cannot be guaranteed the profitability expected at the beginning of the investment. This is called the risk of identification of the stochastic process.