Academic journal article Journal of Money, Credit & Banking

Money Demand and Exchange Rate Determination under Hyperinflation: Conceptual Issues and Evidence from Yugoslavia

Academic journal article Journal of Money, Credit & Banking

Money Demand and Exchange Rate Determination under Hyperinflation: Conceptual Issues and Evidence from Yugoslavia

Article excerpt

THE STANDARD APPROACH to explain exchange rate dynamics under hyperinflation is the Monetary Model of Exchange Rate Determination (MMER) (Frenkel 1976 and MacDonald and Taylor 1992). It states that domestic money supply and demand determine the price level, and purchasing power parity (PPP) subsequently sets the exchange rate. Foreign variables and domestic real variables are ignored since they are completely dominated by domestic monetary factors in hyperinflation.

A recent study of the MMER under rational expectations (RE) for the German hyperinflation in the 1920s (Engsted 1996) gives conflicting results: the exact RE model of the exchange rate holds, while that of the price level does not. These results clearly question the MMER, which states that the exchange rate is determined by prices. Engsted (1996) suggested that the results might be due to severe measurement errors in the price-level series, or, referring to Petrovic and Vujosevic (1995), that the exchange rate is determined directly in the money market, without any reference to prices.(1)

The main purpose of this paper is to explore whether the exchange rate in the Yugoslav hyperinflation was set outright by the money supply and demand, as opposed to the standard MMER explanation. Another way to state this conjecture is that the exchange rate, instead of the price level, should appear in the money demand schedule when measuring real balances. A side issue in this paper is finding a money demand schedule that can explain the whole period of the Yugoslav hyperinflation.

Empirical evidence pointing to the modified MMER advanced in this paper lies in the fact that in hyperinflation the public bases its decisions on the exchange rate rather than on price indices. The exchange rate is a daily observable magnitude set in the market, hence measured without error, which is not the case with price indices. Furthermore, contrary to price indices, the exchange rate is widely followed and well understood. As a consequence, all prices and incomes in hyperinflation are expressed in foreign currency, which becomes the universal unit of account, while only partially the medium of exchange. A natural extension of this is to assume that the public uses the same unit, that is, foreign currency, while determining domestic real money holdings.

There are also some theoretical hints in Cassel (1916) suggesting that relative quantities of money, domestic and foreign, set the exchange rate directly.(2) Likewise, Bresciani-Turroni (1937) advanced the hypothesis that the exchange rate is determined by relative money stocks.

The exchange rate determination under RE can be tested using a log-linear money demand schedule, in particular, Cagan's (1956). This approach gave rise to the present value models of both price level and exchange rate (Engsted 1993, 1996; Obstfeld and Rogoff 1997). If the exchange rate model holds while that of prices does not, our hypothesis, as opposed to the standard MMER, is confirmed. Therefore, we shall first estimate and test these two models.

Second, a nonlinear variable semielasticity money demand schedule will be used. There is already a large amount of literature proposing that Cagan's semielasticity may not always be constant (for example, Bisignano 1975, Frenkel 1976, and Jacobs 1977). Some empirical evidence on the Yugoslav hyperinflation suggests that this might be the case, that is, that the Cagan (1956) constant semielasticity schedule cannot explain the later months of this hyperinflation (Petrovic and Vujosevic 1996, and Engsted 1998). Other evidence also indicate that the Cagan schedule substantially underpredicts the demand for money in the latter stages of some hyperinflations (Cagan 1956; Flood and Graber 1980) hence questioning its validity. There are also theoretical reasons for using a changing semielasticity schedule. Deriving money demand from intertemporal utility maximization, either using the money-inutility function approach (Calvo and Leiderman 1992) or the cash-in-advance approach (Easterly, Mauro, and Schmidt-Hebell 1995), in general gives rise to a variable semielasticity schedule. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.