We examine the so-called "Neo-Fisherian" claim that, at the zero lower bound (ZLB) of the monetary policy interest rate, and the economy in a depression equilibrium, in order to restore the desired inflation rate the policy rate should be raised consistently with the Fisher equation. This claim has been questioned on the ground that the Fisher equation cannot be used mechanically to peg the long-run inflation expectations. It is necessary to examine how inflation expectations are formed in response to, and interact with, policy actions and the evolution of the economy. Hence we study a New Keynesian economy where agents' inflation expectations are based on their correct understanding of the data generations process, and on their probabilistic confidence in the central bank's ability to keep inflation on target, driven by the observed state of the economy. We find that the Neo-Fisherian claim is a theoretical possibility depending on the interplay of a set of parameters and very low levels of agents' confidence. Yet, on the basis of simulations of the model, we may say that this possibility is remote for most commonly found empirical values of the relevant parameters. Moreover, the Neo-Fisherian policy-rate peg is not sustained by the expectations formation process.