Chapter 6 The Monetary Approach To International .

George Alogoskoufis, International Macroeconomics, 2016Chapter 6The Monetary Approach to International MacroeconomicsThe monetary approach is one of the central pillars of international macroeconomics. Its point ofdeparture is the so called monetary model, which identifies the factors affecting long-term nominalexchange rates. The monetary model was originally used as a framework of analysis of the balanceof payments in a fixed exchange rate regime (Frenkel and Johnson 1976), and then as a frameworkfor analysis of the determination of nominal exchange rates in a flexible exchange rate regime(Frenkel and Johnson 1978).The basic monetary approach assumes that there is full flexibility in prices and focuses on theequilibrium conditions in the money market and the international foreign exchange markets.Although this is basically an ad hoc model, like the Mundell-Fleming model, many of its theoreticalproperties are confirmed by inter-temporal optimization models in monetary economies (see Lucas1982).However, the monetary model has a number of empirical shortcomings. It relies on the assumptionof purchasing power parity, which is rejected empirically, and it has difficulties in accounting forthe high volatility of nominal exchange rates observed in practice.The monetary model can be combined with the assumption of gradual adjustment in prices, anddeliver a model that is more in accordance with the evidence, and which, in many ways, looks likethe Dornbusch (1976) model. Thus, combining the monetary model with the assumption of gradualprice adjustment, one can account for phenomena like exchange rate overshooting and the positivecorrelation between nominal and real exchange rates.6.1 Purchasing Power ParityA key component of the monetary approach is the concept of purchasing power parity. The ideaoriginated from the early 19th century, and one can find it in the writings of Ricardo. The idea wasrevived in the early 20th century by Cassel (1921).The approach of Cassel starts with the observation that the exchange rate is the relative price of twocurrencies. Since the purchasing power of the domestic currency is 1/P, where P is the domesticprice level and the purchasing power of the foreign currency is 1/P*, where P* is the foreign pricelevel, the relative price of two currencies should reflect their relative purchasing power. In this case,it should follow that,S P /P*or, in logarithmic form,(6.1)

George Alogoskoufis, International Macroeconomics, 2016s p p*Ch. 6(6.2)where for any variable X, x lnX.It is worth noting that the theory of purchasing power parity can be derived from the IS curve of anopen economy, when the elasticity of aggregate demand with respect to the real exchange rate tendsto infinity. For example, the IS curve in the Mundell Fleming that we examined took the form,! y δ (s p * p) γ y σ i gAs δ tends to infinity, aggregate demand will be finite only if the logarithm of the real exchange rates p*-p tends to zero, that is if (6.2) is satisfied. If the elasticity of aggregate demand with respect tothe real exchange rate is very high, then there cannot be large deviations between domestic andinternational prices, expressed in a common currency, as even small deviations would produce largechanges in aggregate demand. Effectively, the theory of purchasing power parity asserts thatdomestic and foreign goods are perfect substitutes.The purchasing power theory approach essentially requires that the real exchange rate should beconstant. However, this prediction is generally rejected by empirical evidence. Real exchange ratesare not constant, but display considerable fluctuations. Moreover, there seems to be a strongpositive correlation between nominal and real exchange rates, which is not consistent withpurchasing power parity.A variant of this approach, which we will examine below, allows fluctuations in the real exchangerate and treats purchasing power parity as a theory determining the long-term real exchange rate.In any case, for the monetary model with fully flexible prices, the assumption of purchasing powerparity is central.6.2 The Monetary Approach to the Balance of PaymentsThe monetary approach to the balance of payments (Frenkel and Johnson 1976), uses the monetarymodel to explain the behavior of the balance of payments, under a regime of fixed exchange rates.Consider a small open economy that maintains a constant exchange rate through interventions of itscentral bank in the foreign exchange market.The domestic money supply is determined by,! M µ B µ (B f Bd )(6.3)where Μ denotes the money supply, B the monetary base (high powered money), µ the multiplier ofthe monetary base, Bf net foreign exchange reserves of the central bank, and Bd net domestic creditof the central bank to the public and the banking system.In logarithms, assuming a multiplier µ close to unity, (6.3) can be approximated as,!2

George Alogoskoufis, International Macroeconomics, 2016! m θ b f (1 θ )bdCh. 6(6.4)where θ is the equilibrium share of net foreign exchange reserves in the monetary base. (6.4)determines the money supply.The demand for money function takes the form,! m p φ y λi(6.5)where y is the logarithm of domestic output, assumed exogenous, and i the domestic nominalinterest rate.The domestic nominal interest rate is determined by uncovered interest parity. Because of fixedexchange rates, it cannot differ from the international nominal interest rate, assumed exogenous.!i i*(6.6)where i* is the international nominal interest rate.Combining the assumption of fixed exchange rates, uncovered interest parity and purchasing powerparity, by substituting (6.6) and (6.2) in 6.5), we get,m s p * φ y λi *(6.7)where ! s denotes the constant level of the nominal exchange rate.From (6.7), the demand for money is determined by the level at which the exchange rate is fixed(because it co-determines the domestic price level), the international price level, domestic incomeand the international nominal interest rate. All these variables are assumed exogenous.Equilibrium in the domestic money market implies that (6.4) and (6.7) must be satisfiedsimultaneously. Solving for the logarithm of net foreign exchange reserves we get,bf 1 s p * φ y λi * (1 θ )bd θ (6.8)Equation (6.8) incorporates all the predictions of the monetary approach to the balance of payments.A devaluation (increase in s), a rise in the international price level, an increase in domestic outputand income and a reduction in international nominal interest rates, increase the demand for money,and, for given domestic credit, cause increases in foreign reserves. The increase in foreign exchangereserves will occur through surpluses in the balance of payments.!3

George Alogoskoufis, International Macroeconomics, 2016Ch. 6On the other hand, if there is an expansion in domestic credit expansion, the only result will be aloss of net foreign exchange reserves, as the demand for money will not change. Thus, a domesticcredit expansion will cause a deficit in the balance of payments.It is worth noting that the monetary approach to the balance of payments is not concerned with thedetermination of the current account, but the so-called official balance, which is none other than thesum of the current account and the capital account, without taking account of changes in the netforeign exchange reserves of the central bank.6.3 The Monetary Approach to Flexible Exchange RatesIn a regime of flexible exchange rates, the central bank can determine the money supply withoutloss of foreign reserves, such as when it has to intervene in order to stabilize the exchange rate. Thefocus of the monetary approach shifts from the balance of payments to the determination of thenominal exchange rate.The model consists of the money demand function (6.5), the purchasing power parity condition(6.2), and uncovered interest rate parity, which is given by, ! i i * se(6.9) where ! s e is the rational expectation of the change in the exchange rate.Substituting (6.2) and (6.9) in the money demand function (6.5), and solving for the expectedchange in the exchange rate, we have, e!s 1( s m φ y λi * p *)λ(6.10)As the exchange rate is a jump variable, in the absence of bubbles, (6.10) will be satisfied forexpectations of no further depreciation. As a result, the exchange rate will jump to the level,! s m φ y λi * p *(6.11)(6.11) is the basic equation of the determinants of the nominal exchange rate based on the monetaryapproach.Increases in domestic money supply and international interest rates cause a depreciation of thenominal exchange rate, while increases in domestic income and the international price level causean appreciation. In an equilibrium without "bubbles" there can be no expectations of future changesin the exchange rate. The exchange rate, as a non-predetermined variable, immediately adjusts tothe steady state equilibrium described by (6.11).The problem with this model is that it requires that the real exchange rate is constant, and that aslong as domestic inflation differs from international inflation there will be a continuous adjustmentof the exchange rate, in order to satisfy the purchasing power parity condition. However,!4

George Alogoskoufis, International Macroeconomics, 2016Ch. 6empirically, purchasing power parity does not appear to be valid. Real exchange rates fluctuate, andtheir fluctuations are closely related to fluctuations in nominal exchange rates.A variant of the monetary model, combines it with the assumption of the gradual adjustment of theprice level in order to achieve purchasing power parity in the steady state. Thus, the assumption ofpurchasing power parity is only assumed to hold in the steady state and not in the short run.6.4 Gradual Adjustment of the Price Level and the Monetary ApproachSuppose, on the lines of the Dornbusch (1976) model, that the domestic price level adjustsgradually towards its steady state equilibrium level, which is considered to be the price level thatsatisfies purchasing power parity. Instead of the short purchasing power condition (6.2), we nowassume that, ! p π (s p * p)(6.12)where π 0. π is a parameter denoting the speed of adjustment of the domestic price level towardsits steady state equilibrium.The model now consists of the equilibrium condition in the money market (6.5), the uncoveredinterest parity condition (6.9), and the price adjustment equation (6.12).Substituting the uncovered interest parity condition (6.9) in the equilibrium condition for thedomestic money market (6.5), and using the hypothesis of rational expectations, we get, ! s se 1( p m φ y λi *)λ(6.13)(6.13) indicates that as the domestic price level is a predetermined variable, and cannot adjust in theshort run to equilibrate the domestic money market, this role must be played by the domesticnominal interest rate. Since the domestic nominal interest rate can only differ from the internationalnominal interest rate to the extent that there are expectations of future changes in the exchange rate,the expected and actual change in the exchange rate must be such as to maintain equilibrium in thedomestic money market.Our model is now described by (6.12) and (6.13). The equilibrium and the adjustment path arepresented in Figure 6.1, which has significant similarities with the model of Dornbusch (1976), asanalyzed in the previous chapter.The 450 line represents purchasing power parity, as a steady state equilibrium condition. In fact, itrepresents (6.12) for no change in the price level. The vertical line is the long-term equilibriumcondition in the money market (6.13), for a constant exchange rate. The steady state equilibrium isat point E, which is a saddle point, since the price level is a predetermined variable and theexchange rate is a non-predetermined variable. The adjustment path depicted is unique, as all otheradjustment paths lead away from the steady state.!5