In this paper we investigate the possibility that the mathematical conditions are not sufficient to recover the anomalies of the financial rate of return (ROR), when we have multiple roots in the polynomial associated with an investment project. We test empirically the solution in an ex-post agricultural investment cofinanced by EU funds and we find that the financing operations make the correctness of evaluating the ROR questionable. The methodological issues become relevant for policy-maker when the institutions use selective instruments in their decisions. In fact, the cut off used in the International Institutions might not be optimal to decide the project rank.