Monetary hyperinflations, speculative hyperinflations and modelling the use of money.

The aim of this paper is to clarify the failure of the Cagan model with perfect foresight and to draw new axes for investigation of monetary hyperinflation analysis. Firstly, the paper evaluates the relevancy of the Cagan ad-hoc model with perfect foresight as a theoretical framework for investigating hyperinflation processes. We show that deficits can never generate monetary hyperinflations, confirming the results of Buiter (1987). The only hyperinflationary processes that can occur are speculative hyperinflations. Secondly, the paper assesses consistency of hyperinflationary paths with the optimizing behaviour of representative agents within two perfect foresight inflationary finance frameworks modelling the use of money as a medium of exchange. In the context of a money-in-the-utility framework, the results obtained in the Cagan ad-hoc model with perfect foresight are founded and confirmed. This implies restricting the use of the latter model only to speculative hyperinflations analysis. In the context of a transaction costs based model, we show that deficits can generate monetary hyperinflations. Moreover, speculative hyperinflations remain possible. This result is in sharp contrast to that of the money-in-the-utility framework and implies a demand for money different from the Cagan form.