"Money demand in the Yugoslavian hyperinflation 1991-1994"  
	
Bent Nielsen 
Nuffield College, University of Oxford

Abstract:	
Empirical analyses of Cagan's money demand schedule have broadly
speaking suffered from the following problems: (i) Inability to model
the data to the end of the hyperinflation. (ii) Difficulties in making
congruent models for systems of variables. (iii) Discrepancies between
"estimated" and "actual" inflation tax.  In this paper the extreme
Yugoslavian hyperinflation of the 1990's is therefore studied.  Two
econometric models are presented.  First, money, prices and exchange
rates are analysed by a vector autoregression allowing for random walk
and explosive common trends.  This analysis of the sample distribution
leads on to the second model of real money and the cost of holding
money, rather than the traditional inflation measure.  The three
outlined problems can then be addressed, giving support to Cagan's
model.