For exchange-rate expectations, the simplest assumption is that in-\ estors do not expect the exchange rate to change. This assumption can be justified both on the grounds of ease and on the grounds that it is difficult to find evidence of predictable exchange-rate movements (Meese and Ro-goff, 1983). These assumptions about capital mobility and exchange-rate expectations lead to the famous Mundell-Fleming model (Mundell, 1968; Fleming, 1962).

Perfect capital mobility implies that if there were any difference in the expected rate of return between domestic and foreign assets, investors would put all of their wealth into the asset with the higher yield. Since both types of assets must be held by someone, it follows that the expected rates of return on the two assets must be equal. The expected rate of return on foreign assets in terms of domestic currency is the foreign mterest rate plus any expected increase in the price of foreign currency. With static exchange-rate expectations, the expected change in the price of foreign currency is zero. Thus the requirement that the expected rates of return are equal is simply

i =i*, . • (5.13)

where i * is the foreign interest rate; j * is taken as given.

At this point it is necessary to distinguish between floating and fixed exchange rates. With a floating exchange rate, aggregate demand is described by the three equations (5.1), (5.12), and (5.13) in the three unknowns i, , and 8. Since z is determined trivially by the requirement that It equals /*, the system immediately reduces to two equations in Y and f:

-=L(i*,Y), (5.14)

Y = E(Y,i* -Tr,G,T,8P*/P). (5.15)

Figure 5.6 plots the sets of points satisfying (5.14) and (5.15) in output-exchange rate space. Since an increase in fP* / raises planned expenditure, the set of solutions to (5.15) is upward-sloping; this is shown as the /5 * curve in the figure. And since the exchange rate does not affect money demand, the set of solutions to (5.14) is vertical; this is shown as the IM* curve.

The fact that the LM* curve is vertical means that output for a given price level-that is, the position of the AD curve-is determined entirely in the money market. To take the same example as in the previous section, suppose that government purchases rise. This change shifts the IS* curve to the right. As shown in Figure 5.7, however, at a given price level this leads only to appreciation of the exchange rate and has no effect on output. Thus the aggregate demand curve is unaffected.

Assuming a fixed rather than a floating exchange rate requires two changes to the model. Fhst, the exchange rate is now pegged at some level 1:

e = £. (5.16)

FIGURE 5.6 The Mundell-Fleming model with a floating exchange rate

FIGURE 5.7 The effects of an increase in government purchases with a floating exchange rate

FIGURE 5.8 The Mundell-Fleming model with a fixed exchange rate

Second, the money supply becomes endogenous rather than exogenous. For the government to fix the exchange rate, it must stand ready to buy or sell domestic currency in exchange for foreign currency at the rate ?. It therefore cannot independently set M, but must let it adjust to ensure that the exchange rate remains at I.

The aggregate demand side of the model with a fixed exchange rate therefore consists of the LM equation, (5.1); the IS equation, (5.12); the interest-rate equation, (5.13); and the exchange-rate equation, (5.16). Once again, we can substitute the i = i* condition into the /5 and LM equations to simplify the system. This gives us the LM* equation, (5.14); the /5* equation, (5.15); and the exchange-rate equation, (5.16). In addition, the LM* equation, M/P = L{i*, Y), serves only to determine M, and can therefore be neglected. Thus we are left with the IS * equation and the exchange-rate equation. The /5 * curve is upward-sloping as before; and the exchange-rate equation is simply a horizontal line at ?. Figure 5.8 depicts the solutions to these equations in output-exchange rate space.

The results for this case are the opposite of those for a floating exchange rate. Changes in planned expenditure now affect aggregate demand. A rise in government purchases, for example, shifts the /5* curve to the right and thus raises output for a given price level. Disturbances in the money market, in contrast, have no effect on Y for a given P. A rise in the demand for money, for example, leads only to an increase in the money supply.