Should we expect the large US trade deficit and current account deficit to disappear naturally in the future? At an unchanged real exchange rate, the answer is 'Probably not.
If there were good reasons to expect US trading partners to experience much higher growth than the United States over the coming decade, then we could expect to see the same process we saw in the 1990s, but this time in reverse. Lower growth in the United States than in the rest of the world would lead to a steady reduction in the trade deficit.There are few reasons, however, to expect such a scenario. All countries are expected to slow down in 2008-09. but there is no reason to expect the United States to grow more slowly than Europe and Japan. Put another way, there is no particular reason to expect that the trade deficit will narrow by itself, without a further depreciation of the dollar.
Can we expect the shifts in exports and imports to reverse themselves, leading to an improvement in the trade balance without the need for a depreciation? The source of the shifts is poorly understood, so one must be careful in predicting what might happen. But there doesn't appear to be any particular reason to think that, for example, US consumers will shift back from foreign cars to US cars.
Will the dollar depreciate further, leading eventually to a reduction in the trade and current account deficits? The answer is'Probably'. While financial investors have been willing to lend to the United States until now, there are many reasons to think that they will be reluctant to lend at the current levels of $800 billion or so per year. (More on this in the focus box 'Sudden stops, the strong US dollar, the vulnerable Australian dollar and the limits to the interest parity condition' in Chapter 20.) These arguments have three implications:
• The US trade and current account deficits will decline in the future.
• This decline is unlikely to happen without a further real depreciation. How large a depreciation? Estimates range from 10 per cent to 30 per cent from where we were in 2008—in short, a substantial further real depreciation.
• When will this depreciation take place? This is a hard question to answer. It will take place when foreign investors become reluctant to lend to the United States at the rate of $800 billion or so per year. Let's go back to the issues discussed in Table 19.1. A depreciation on such a scale will have major effects
on the demand for goods both in the United States and abroad.
The depreciation will increase the demand for US goods. When the depreciation takes place, if US output is already close to its natural level, the risk is that the depreciation will lead to too high a level of demand and too high a level of output If this happens, it will require a decrease in domestic demand.This may come from a decrease in spending by consumers or firms or from a reduction in government spending. If the US government succeeds in achieving a smooth depreciation and the decrease in domestic spending, the outcome can be sustained US growth and a reduction of the US trade deficit.
The depreciation will decrease the demand for foreign goods. By the same argument, this may require foreign governments to implement policies to sustain their own demand and output.This would ordinarily call for a fiscal expansion, but t might not be the right solution in this case. A number of countries, such as France and Japan, are already rurning large budget deficits. For the reasons we saw in Chapter 17, increasing these deficits further may be d fficult and even dangerous. If fiscal policy cannot be used to sustain demand and output, a large dollar depreciation might therefore trigger a recession in those countries.
1
In short, a smooth reduction of the US trade deficit will require the combination of a dollar depreciation and spending changes both in the United States and abroad.This can be achieved, but it may not be easy.The global financial crisis certainly doesn't appear to be making it easier.
19.5 LOOKING AT DYNAMICS:THE J-CURVE
Wc have ignored dynamics so lar in this chaptcr. It is time to reintroduce them. The dynamics of consumption, investment, sales and production discussed in Chapter 3 are as relevant to thc open economy as they are lo the closed economy. But there are additional dynamic eflects as well, which come from lhe dynamics ot exports and imports. We focus on these effects here.
Return to the effects ol the exchange rate on the trade balance. We argued earlier that a depreciation leads to an increase in exports and to a decrease in imports. Bui these effects don't happen overnight. Think ol the dynamic ellects ol. say, a 10 per cent Australian dollar depreciation.
In the lirst few months following the depreciation, the ellect ol the depreciation is likely to be reflected much more in prices than in quantities. The price ol imports in Australia goes up, the price of Australian exports abroad goes down. But the quantity ot imports and exports is likely to adjust slowly. It lakes a while tor consumers to realise that relative prices have changed, it takes a while for firms to shilt lo cheaper suppliers, and so on. So, a depreciation may well lead to an initial deterioration of thc trade balance,- e decreases, but neither X nor IM adjusts very much initially, leading to a decline in net exports (X - IM/e).
As time passes, the clfects ol the change in the relative prices ol both exports and imports become stronger. Cheaper Australian goods lead Australian consumers and lirms to decrease their demand for foreign goods: Australian imports decrease. Cheaper Australian goods abroad lead foreign consumers and firms to increase their demand for Australian goods: Australian exports increase. If the Marshall I.erner condition eventually holds—and we have argued that it does—the response of exports and imports eventually becomes stronger than the adverse price effect, and lhe eventual effect of the depreciation is to improve the trade balance.
The response of the ► trade balance to the real exchange rate: Initially: (X,/M) unchanged, e I =>(X- IMIe) I Eventually: (X tlMl.el) => (X - IMIe) T.
Figure 19.6 captures this adjustment by plotting the evolution ol the trade balance against time in response to a real depreciation. The pre-depreciation trade delicit is OA. Thc depreciation initially increases the trade delicit to OB: e goes down, but neither IM nor X changes right away. Over time, exports increase and imports decrease, reducing the trade deficit. Eventually (if the Marshall-Lerner condition is satisfied), the trade balance improves beyond its initial level,- this is what happens from point С on in the ligure. Economists refer lo this adjustment process as the J-curve because— admittedly, with a bit ol imagination—the curve in the ligure resembles a /: first down, then up.
Figure 19.6 The J-curve
The importance ol the dynamic ellects ol the real exchange rate on the trade balance can be seen from the evidence from Ausiralia since 1980. Figure 19.7 plots thc Australian trade deficit (as a proportion of CDP' against lhe Australian real multilateral exchange rate (lagged by one year) since the
Depreciation
I Time
О
A ж C
81
A real depreciation leads initially to a deterioration of the trade balance, then to an improvement
1980s. They do tend to move together in a statistically significant way. At the beginning of the 1980s, before Australian linancial markets and capital movements had been liberalised thc real exchange rate- had increased to too strong a level. As a consequence, the trade deficit increased, reaching a high value ol nearly 4 per ccnt ol GDP in 1982. After deregulation in late 1983, the nominal Australian dollar depreciated sharply, falling 3 I per ccnt in two years to 1986. Within government and financial circles, panic set in- epitomised by the notorious banana republic' speech by the treasurer, Paul Keating. However, with the benefit of hindsight, wc see in Figure 19.7 how that real depreciation encouraged a subsequent gradual i and variable) correction ol the trade deficit, expressed as a proportion of GDP. Two facts arc clear:
1. Movements in the real exchange rate were reflected in parallel movements in the trade deficit. The appreciation was associated with a large deterioration ol thc trade balance, and the later deprecia¬tion was associated with a large improvement in the trade balance. From 2001 to 2005, the real exchange rale appreciated by 17 per cent,- accordingly, the trade deficit widened to reach 2.5 per cent ol GDP- From 2006-08, the real exchange rale continued appreciating (until mid-2008), but the trade delicit improved. Why was this? The reason is that global commodity prices increased, which enhanced export revenues.
2. There were substantial lags in the response of the trade balance to changes in ihe real exchange rate. Note how from 1984 to 1986 the trade deficit worsened even while the dollar was depreciating so much, but improved from 1987. And note how the steady depreciation of ihe real Australian dollar Irom 1997 to 1999 saw an initial worsening of the trade deficit ilrom a small surplus in 1997 to a 2.4 per cent delicii in 1999) and then a substantial improvement thereafter. Thc dynamics ot thc J-curve were very much ai work in both episodes.
In general, the econometric evidence on the dynamic relation between exports, imports and the real exchange rate suggests that in all OECD countries a real depreciation eventually leads to a trade balance improvement. But it also suggests that this process takes some lime, typically between six months and a year. These lags have implications not only for the effects of a depreciation on thc trade balance but also for the effects ot a depreciation on output. If a depreciation initially decreases: net exports, it also initially exerts a contractionary effect on output. Thus, il a government relies on a depreciation both to improve the trade balance and to expand domestic output, the effects will go the 'wrong way for a while.
Figure 19.7
The real
Д
ь exchange rate
rr
o' and the ratio of
О the trade deficit
fT
3 to GDP: Australia,
a
n 1980-2008
a
=> n
n
r*
О
Ci
О
"0
•—
I n
90
Lagged real ^exchange rate (scale at left)
1980
I I I I I I I
1984 1988
Trade deficit/GDP (scale at right)
Г 4
i I I I I I I
1992 1996
I l I I I I I I I
2000 2004 2008
о ®
о
и м с
я
и
X
о
"а 6 К
The very strong real Australian dollar in the early 1980s was associated with a large trade deficit The depreciation since 1984 led to an improvement in the trade deficit The strong appreciation since 2001 has led lo a large trade deficit There were, however, substantial lags in the effects of the real exchange rate on the trade deficit
SOURCES: IMF International Financial Statistics: RBA BulleCn. Table GI0.
19.6 SAVING, INVESTMENT AND THE CURRENT ACCOUNT BALANCE
You saw in Chaptcr 3 how we could rewrite thc condition lor equilibrium in the goods market as the condition that investment equals saving—the sum ol private saving and public saving. Let lis now derive the corresponding condition lor the open economy, and show how uselul this alternative way of looking at the equilibrium can be.
Start Irom our equilibrium condition:
Y = С + / + С - IM/e + X
In Chapter IS we explained that the current account (С A i sums up the trade account IM/e - Xi and the net investment and translers account I which we will define as NIA). Also we explained that CNP is simply the sum ol NIA and CDP.
So, adding NIA to both sides of the above equation gives
Y + NIA С + 1 + G- IM/e + X + NIA
or
GNP = С + / + G + CA
Subtract С + Г from both sides, and use the lact that private saving is given by S = GNP - С - I to get
5 - I + G - T - CA
Reorganising gives
CA = S + IT - G) - I (19.5)
This condition says that, in equilibrium, the currcnt account, СЛ. must equal saving (private saving, -V and public saving T - G) minus investment. 1. It follows that a current account surplus must correspond to an excess ol saving over investment; a current account delicil must correspond to an excess ol investment over saving.
One way of getting more intuition lor this relation is to return lo thc discussion ol the current account and the capital account in Chapter 18. There, we saw that a current account surplus implies an equal capital account delicit or net lending Irom the country lo lhe resl ol the world, and a current account delicit implies net borrowing hy the country from the resl of the world. So. consider a country that invests more than it saves so that 5 + (T-G) - I is negative. That country must be borrowing the difference from the rest ol the world, it must therefore be running a current account deficit. Nole some ol the things that equation 119.5) says:
• An increase in investment must he reflected either in an increase in private saving or public saving or in a deterioration ot the current account balance 'a smaller current account surplus, or a larger current account delicit).
• An increase in the budget delicil must be reflected in an increase in private saving, or a decrease in investment, or a deterioration ot ihe current account balance.
• A country with a high saving rate private and public must have either a high investment rate or a large airreni account surplus.
Note also, however what equation (19.51 does not
The argument may sound convincing but we know il is wrong. Let's assume thai the net investment and translers account i SIA remains relatively constant. We showed earlier that a depreciation leads to nn increase in output and an improvement in the trade account position. So, what is wrong with the argument? A depreciation actually allects saving and investment II does so by affecting the demand lor domestic goods, thereby increasing output. Higher output leads to an increase in saving over investment, or equivalently to a decrease in thc trade deficit.
Л good way of making sure that vou understand the material in this chapter is to go back and look ai the various cases we have considered, Irom changes in government spending, to changes in loreign output, to combinations of depreciation and liscal contraction, and so on. 'Irace what happens in each case to each of the lour components ol equation 1 19.51: private saving, public saving (equivalently, the budget surplus), investment and the trade balance. Make sure as always, that you can tell the story in words. If you can, you arc ready to go on to Chapter 20.
SUMMARY
In an open cconomy, the demand lor domestic goods is equal to thc domestic demand tor goods (consumption, plus investment, plus government spending! minus the value ot imports 'in terms ot domestic goods), plus exports.
In an open economy, an increase in domestic demand leads to a smaller increase in output than it would in a closed economy because some ot the additional demand tails on imports. For the same reason, an increase in domestic demand also leads to a deterioration of the trade balance. An increase in foreign demand leads as a result of increased exports, to both an increase in domestic output and an improvement in the trade balance.
Because increases in foreign demand improve the trade balance and increases in domestic demand worsen the trade balance, countries may be tempted to wait tor increases in loreign demand lo move them out ot a recession. When a group ot countries is in recession, coordination can help them get out of it.
II the Marshall-Lerner condition is satisfied—and thc empirical evidence suggests lhal it is—a real depreciation leads to an improvement in net exports.
A real depreciation leads first to a deterioration ot thc trade balance and then to an improvement. This adjustment process is known as the l-curvc.
Show, for example, that an increase in foreign demand leads to:
• an increase in private saving
• an increase in investment (but by less than private saving)
• no change in the budget deficit
• an improvement in 1 the trade balance.
Thc condition lor equilibrium in the goods market can be rewritten as ihe condition that saving public and private) minus investment must be equal to the currcnt account balance. A current account surplus corresponds to an cxcess of saving over investment. A currcnt account dcticit corresponds to an excess of investment over saving.
G-8. C.-20, ASFAN. 438 Marshall-Lerner condition. 440 l-curvc, 446
QUESTIONS AND PROBLEMS
Quick check
I. Using the information in this chapter, label each of the following statements 'true', 'false' or
'uncertain'. Explain briefly.
a. Trade deficits generally reflect high investment, not low national saving.
b. The national income identity implies that budget deficits cause trade deficits.
c. Opening the economy to trade tends to increase the multiplier because an increase in expendi¬ture leads to more exports.
d. The only way a country can eliminate a trade surplus is through a real appreciation.
e. A small open economy can reduce its trade deficit through liscal contraction at a smaller cost in output than a large economy can.
f. If thc trade deficit is equal to zero, thc domestic demand for goods and the demand lor domestic goods are equal.
g. Thc current high LIS trade deficit is the result of higher growth in the Llnited Stales than in ihe rest ol the world since the mid- 1990s.
1. Real and nominal exchange rate and inflation
Using the definition of the real exchange rate tand propositions 7 and S in Appendix 2 at the end of
the book), you can show that the following is true:
Ae A E AP AP~ T = E + P ~ P'
In words, the percentage real appreciation equals the percentage nominal appreciation plus the
difference between domestic and foreign inflation.
a. If domestic inflation is higher than foreign inflation, hut the domestic country has a fixed exchange rate, what happens to the real exchange rate over time? Assume that thc Marshall-Lerner condition holds. What happens to the trade balance over time? Explain in words.
b. Suppose the real exchange rate is constant, say ai the level required for net exports for the current account) to equal zero. In this case, if domestic inflation is higher than foreign inflation, what must happen to the nominal exchange rate over time?
3. The potential effects of a recession in japan on the Australian economy
a. The share ol Japanese spending on Australian goods is 20 per cent of Australian exports, which are themselves equal to about 2.3 per cent of Australian CDP. What is the share of Japanese spending on Australian goods relative to Australian CDP?
b. Assume that the multiplier in Australia is 2. and that a recession in lapan has reduced output by 5 per cent (relative to its natural level). What is the impact on Australian GDP of thc Japanese- slowdown?
c. Il the Japanese recession also leads to a slowdown of thc other economies that import goods Irom Australia, the elfect could be larger. Assume that Australian exports fall by 5 per cent (ol themselves). What is the impact ol this on Australian GDP?
d. Comment on the following statement Irom an economist on television: 'Unless Japan recovers from recession quickly, growth will grind to a halt in the rest of the world.
4. Briefly explain in words the results in Table 19.1.
Dig deeper
5. Net exports and foreign demand
a. Suppose that there is an increase in loreign output. Show the ellect on the domestic economy that is replicate Figure 19.4). What is the effect on domestic output? On domestic net exports?
h. If the interest rate remains constant, what will happen to domestic investment? If taxes arc fixed, what will happen to thc domestic budget dciicit?
c. Using equation ( 19.5), what must happen to private saving? F.xplain.
d. Foreign output docs not appear in equation (19.5), yet it evidently affects net exports. F.xplain how this is possible.
6. Eliminating a trade deficit
a. Consider an economy with a trade deficit (NX < 0) and with output equal to its natural level. Suppose that, even though output may deviate from its natural level in the short run, it returns to its natural level in the medium mn. Assume that the natural level is unaffected by thc real exchange rate. Whai must happen to the real exchange rate over the medium run to eliminate the trade deficit (that is, to increase \'X to 0)?
b. Now write down the national income identity. Assume again that output returns to its natural level in the medium mn. If NX increases to 0 what must happen to domestic demand in the medium run? What government policies arc available to reduce domestic demand in thc medium run? Identity which components of domestic demand each ol these policies affect.
7. Multipliers, openness and fiscal policy
Consider an open economy characterised by the equations below:
С = c() + f| fY-T) I = d„+ d, Y IM = ш, V X = .г, У
The parameters m, and x7 are the propensities to import and export. Assume that the real exchange rate is fixed at a value of J and treat foreign income, Y\ as fixed. Also assume that taxes are fixed and that government purchases are exogenous I that is, decided by the government). We explore the effectiveness of changes in С under alternative assumptions about the propensity to import.
a. Write thc equilibrium condition in thc market tor domestic goods and solve lor Y.
b. Suppose that government purchases increase by one unit. What is the effect on output? (Assume that 0 < ж, < C| 4- (/, < I. Explain why.)
c. How do net exports change when government purchases increase by one unit?
Now consider two economies, one with m, = 0.5 and the other with m; = 0.7. Each economy is characterised by Ic( i d.) = 0.6.
d. Suppose that one o: the economies is much larger than the other. Which cconomy do you expect to have thc larger value ol m,? Explain.
c. Calculate your answers to parts (bi and c) for each economy by substituting thc appropriate parameter values.
I In which economy will fiscal policy have a larger ellect on output? In which economy will fiscal policy have a larger effect on net exports?
8. Policy coordination and the world economy
Consider the following open economy. The real exchange rate is fixed and equal to one. Consumption, investment, government spending and taxes are given by
С = 10 + 0.8(Y - T); I = Ю; С = 10; T - 10
Imports ami exports are given by
IM - 0.3V; X = 0.3Y* where an asterisk denotes a foreign variable.
a. Solve lor equilibrium output in the domestic economy, given Y*. What is thc multiplier in this economy? II we were to close the economy—so exports and imports were identically equal to zero—what would the multiplier be? Why are they diUcrent?
b. Assume lhat the foreign economy has the same equations as the domestic economy (with asterisks reversed). Use the two sets ol equations to solve lor the equilibrium output ol each country. What is the multiplier lor each country now? Why is it different Irom the open economy multiplier above?
c. Assume thai both countries have a target level of output of 125. What is the increase in G necessary in either ol these countries, assuming that the other country doesn't change its level of government spending, to achieve target output? Solve for net exports and the budget deficit in each country.
d. What is the common increase in G necessary to achieve target output in both countries?
e. Why is liscal coordination such as the common increase in G in [d]> difficult to achieve in practice?
Explore further
9. Get from your library a recent issue of the International Financial Statistics, published monthly by the IMF /also available on CD-ROM or on-line from the IMF,
10. The Australian trade deficit and investment
a. l.ook at equation 19.51. Define national saving as private saving plus the government surplus— that is, as S-T-G.
With this definition in mind, by accounting, how is an increase in the trade deficit related to the difference between national saving and domestic investment?
Co the statistical tables on the RBA website lwww.rba.gov.au) and download Table GT1. Construct annual data for GDP, domestic investment lall types) and net exports from 1980 to the most recent year available. ISet exports are obtained by subtracting imports from exports for each pear.) Divide domestic investment and net exports by GDP for each year to express their values as a percentage of GDP.
b. When did Australia last have a trade surplus?
Subtract the value of net exports las a percentage of GDP/ in that year from the value of net exports las e. percentage of GDP) in the most recent year available. Do the same for domestic investment. Has the decline in net exports been matched by an equivalent increase in investment? What do your calculations imply about lhe change in national saving between then and the present?
c. Is a trade dclicit more worrisome when not accompanied by a corresponding increase in investment? Hxplain your answer.
We invite you to visit the Blanchard-Sheen page on the Pearson Australia website at
www.pearson.com.au/highered/blanchardsheen3e
for many World Wide Web exercises relating to issues similar to those in this chapter.
FURTHER READINGS
• A good discussion ol thc relation between trade deficits, budget delicits, private saving and investment is given in Barry Bosworth's Saving and Investment in a Global Economy (Washington, DC.: Brookings Institution 1993).
• A good discussion of the US trade deficit and its implications for the future is given in William Cline The United States as a Debtor Nation (Washington, DC: Peterson Institute, 2005).
• Read the speech by the rormer RBA Governor Ian Macfarlane on 'What are the global imbalances?' and Recent trends in world saving and investment patterns', in Reserve Bank of Australia Bulletin. October 2005,
APPENDIX: DERIVATION OF THE MARSHALL-LERNER CONDITION
Start from the definition of net exports. NX =X- IMIe. and assume trade to be initially balanced,so that X = IMIe. The Marshall-Lerner condition is the condition under which a real appreciation, an increase in e, leads to a decrease in net exports.
To derive this condition, consider an increase in the real exchange rate of Ae.The change in the trade balance is given by
ANX = AX - (.MM)lt + (/M/e)(Де/е)
The first term on the right, AX. gives the change in exports. The second, (AIM)le, is equal to the change in the quantity of imports in terms o1 domestic goods valued at the original real exchange rate.The third, (7Л1/е)(Ае/е), is equal to the quantity of imports in terms of domestic goods times the rate of change in the real exchange rate of domestic for foreign goods.
Divide both sides of the equation by X to get
ANX ЛХ ДIM + IM Ae X X eX eX e
Use the fact that IMIe = X to replace eX by IM in the second term on the right, and to replace IMleX by 1 in the third term on the right-This substitution gives
ANX _ AX Л//И д6 X X " IM + ~
The change in the trade balance (as a ratio to exports) in response to a real appreciation is equal to the sum of three terms:
• The first term is the proportional fall in the volume of exports, ЛXIX. induced by the real appreciation.
• The second term is equal to minus the proportional rise in the volume of imports. -А//И//М, induced by the real depreciation.
• The third term is equal to the proportional change in the real exchange rate, Ae/e, or, equivalently. the rate of real appreciation.
The Marshall-Lerner condition is the condition that the sum of these three terms be positive. If it is satisfied, a real appreciation leads to a worsening in the trade balance; a real depreciation leads to an improvement in the trade balance.
A numerical example will help here. Suppose that a 1 per cent depreciation leads to a relative increase in exports of 0.9 per cent and to a relative decrease in imports of 0.8 per cent. (Econometric evidence on the response of exports and imports to the real exchange rates suggests that these are indeed reasonable numbers.) In that case, the right-hand side of the equation is equal to 0.9% - (-0.8%) - 1% = 0.7%. Thus, the trade balance improves: the Marshall-Lerner condition is satisfied.
CHAPTER
Output, the Interest Rate and the Exchange Rate
I
n Chapter 19 we treated the exchange rate as if it was one of the policy instruments available to the government. But the exchange rate isn't a policy instrument. Rather, it is determined in the foreign- exchange market—a market where, as you saw in Chapter 18, there is an enormous amount of trading. This fact raises two obvious questions: What determines the exchange rate? How can policy-makers affect it?
These are the questions that motivate this chapter. More generally, we examine the implications of equilibrium in both the goods market and financial markets, including the foreign-exchange market. This allows us to characterise the joint movements of output, the interest rate and the exchange rate in an open economy. The model we develop is an extension to the open economy of the IS-LM model you saw in Chapter 5. and is known as the Mundell-Fleming model, after the two economists. Robert Mundell and Marcus Fleming, who first put it together in the 1960s. (Just as we modified the standard IS-LM model to cater for contemporary economies, the model presented here keeps the spirit of the original Mundell-Fleming model but differs in its details.)
• Section 20.1 looks at equilibrium in the goods market.
• Section 20.2 looks at equilibrium in financial markets, including the foreign-exchange market.
• Section 20.3 puts the two equilibrium conditions together and looks at the determination of output, the interest rate and the exchange rate.
• Section 20.4 looks at the role of policy under flexible exchange rates.
• Section 20.5 looks at the role of policy under fixed exchange rates.
20.1 EQUILIBRIUM INTHE GOODS MARKET
Equilibrium in thc goods market was thc focus of Chapter 19, where we derived the equilibrium condition (equation [19.4]):
Y = C(Y - T) + KY,r) + G - lM(Y,e)/e + X(Y*,e) ( + ) (+,-) (+,+) (+,-)
For the goods market to be in equilibrium, output (the left side ol the equation) must be equal lo the demand for domestic goods l the right side of thc equation). 4
equilibrium (IS):
This demand is equal to consumption, C, plus investment, /, plus government spending, G, minus Output = Demand for imports, IM/e plus exports. X. domestic goods.
®
Consumption. С depends positively on disposable incomc. Y- T.
Investment. /. depends positively on output, V, and negatively on the real interest rate, r.
Government spending, C, is taken as given.
The volume ol imports, IM. depends positively on output, V, and positively on thc real exchange rate, e.
P.xports. X, depend positively on foreign output, V*. and negatively on the real exchange rate, e.
It will be convenient in what follows to regroup the last two terms under net exports', defined as exports minus imports, X IM/e-.
NX(Y,Y*,e) = X(Y*,e) - IM(Y,e)/e
It follows trom our assumptions about imports and exports that net exports, NX depend on domestic output, Y, foreign output, V*, and the exchange rate e. An increase in domestic output increases imports, thus decreasing net exports. An increase in loreign output increases exports, thus increasing net exports. An increase in e— a real appreciation—leads to a decrease in net exports.
NX(Y,Y*,e)
I(Y.r)
(20.1)
First simplification: t P = P* = 1, so, e = E
Second simplification: ► if - 0. so, r = i.
Using this definition of net exports, we can rewrite the equilibrium condition as
Y = C(Y-T) ( + )
For our purposes, thc main implication of equation (20.1) is thc dependence ot demand, and so ol equilibrium output, on both the real interest rate and the real exchange rate:
• An increase in the real interest rate leads to a decrease in investment spending, and so to a decrease in thc demand tor domestic goods. This leads, through the multiplier, to a dccrcasc in output.
• An increase in the real exchange rate—a real appreciation—leads to a shift in demand away from domestic goods, and so to a dccrcasc in net exports. Thc dccrcasc in net exports decreases thc demand for domestic goods and so decreases output.
For :hc remainder of the chapter, we shall make two simplifications to equation (20.1):
• Given our focus on the short run, we assumed in our previous treatment of the IS-LM model that thc domestic) pricc level was given. Wc will extend this assumption to thc foreign pricc level, so thc real exchange rate (e= F.PIP* and the nominal exchange rate (E) move together. A nominal depreciation leads, one-for-one, to a real depreciation. If, lor notational convenience we choose P and P* so that PIP* I (and wc can do so bccausc they arc index numbers), then e E and wc can replace e by £ in equation (20.1 ).
• As we take thc domestic pricc level as given and fixed there is no inflation, neither actual nor expected. The nominal interest rate and the real interest rate are the same, and we can replace thc real interest rate, r, in equation (20.1 by thc nominal interest rate, /.
Reminder: A real appreciation is represented by an increase in the real exchange rate—an increase in the price of domestic goods in terms of foreign goods. ►
We will assume ► throughout the chapter that the Marshall-Lerner condition—that is. the condition that a real appreciation worsens the trade balance— holds (see Chapter 19).
With these two simplifications, equation (20.1) becomes
У = C(Y - T) ( * )
I(Y,i) M
NX(YY*,E)
(20.2)
We leave aside the other choices—between short-term and long- term bonds, and between short-term bonds and stocks— studied in Chapter 15. ► consider each choice in turn.
Output depends on both the nominal interest rate and the nominal exchange rate.
20.2 EQUILIBRIUM IN FINANCIAL MARKETS
When we looked at financial markets in thc IS-LM model, we assumed that people chose between only two financial assets, money and bonds. Now that we look at a financially open economy, we must take into account thc fact that people have a choicc between domestic bonds and foreign bonds. Let's
Money versus bonds
When looking at financial market equilibrium in the model we wrote the condition that the
supply of money be equal to the demand lor money as
YL i
Two qualifications from Chapter 18:(I) US dollars used for illegal transactions abroad; and (2) US dollars used for domestic transactions in countries with very high inflation. We will ignore 4 both qualifications here.
4 Financial markets equilibrium. Condition (LM): Supply of money : Demand for money.
4 In Secdon 20.5 we will introduce fixed exchange rate regimes. For countries that select these, their monetary policy is directed at the exchange rate rather than the interest rate.
M P
< 20.3)
We look the price level, P as given. We assumed that the real demand for money (the right side of equation [20.3]) depended on the level ol transactions in the economy, measured by real output, Y, and on the opportunity cost of holding money rather than bonds, lhe nominal interest rate on bonds, /'.
1 low should wc change this characterisation now that thc economy is open? You will like the answer: Not very much, il at all.
In an open economy, the demand for domestic money is still mostly a demand by domestic residents. There isn't much reason for, say. Americans to hold Australian currency or Australian dollar- denominated demand deposits. Transactions in the United States require payment in US dollars, not in Australian dollars. II residents ol the United States want to hold Australian-dollar-denominatcd assets, they are better oil holding Australian bonds which at least pay a positive interest rate. And the demand lor money by domestic residents in any country still depends on the same factors as belore: their level ol transactions that we measure by domestic real output, and the opportunity cost of holding money, thc nominal interest rate on bonds.
Therefore, we can still use equation 20.3) to think about domestic money market equilibrium in an open economy. The supply ol domestic money must equal the demand lor domestic money in equilibrium.
In our closed economy analysis in Chapter 5. we assumed that the interest rate was determined as a policy choice of the central bank, which meant that the money supply VI. was endogenous. The central bank could achieve its desired interest rale by undertaking lhe appropriate opcn-markcl operations in money and domestic bonds. This characterisation will apply as long as thc central bank doesnt want to achieve a desired exchange rale. Or putting it another way, our core analysis of interest rate setting based on inflation targeting will work in a regime with flexible exchange rates. This extended analysis applies to Australia today and many other countries.
As wc did in Chapter 5. monetary policy is characterised as fixing the interest rate at some value, i0:
Interest rate setting- / = /„
In Chapter 21, Section 21.3, we will explain what happens il the central bank uses a more sophisti¬cated monetary policy rule with a pricc target as you saw in Chapter 7, Section 7.2).
Domestic bonds versus foreign bonds
In looking at the choice between domestic bonds and foreign bonds, we will rely on the assumption introduced in Chapter 18: linanciai investors, domestic or foreign go lor the highest expected rate of return. This implies that in equilibrium domestic bonds and foreign bonds must have thc same expected rate of return,- otherwise, investors would be willing to hold only one or the other, but not both and this couldn't he an equilibrium. Like most other economic relations, this relation is only an approximation to reality and does not always hold. (More on this in the locus box Sudden stops, thc strong LIS dollar, the vulnerable Australian dollar and limits to the interest parity condition at the end ol this section.
The assumption that domestic and loreign bonds have the same expected return is sometimes described as perfect capital substitutability. Thc relative riskiness ol investing in domestic or foreign bonds is assumed to be irrelevant. It isn't difficult to introduce imperfect capital substitutability, but wc leave that to an appendix at the end ot this chapter. We will also continue to assume that there are no international capital controls, or that there is perfect capital mobility. Again we deal with issues ol imperfect capital mobility in an appendix.
As you saw in Chapter 18. equation (18.2), this assumption implies that the following arbitrage relation—the exact interest parity condition—-must hold:
. E,
1 + i, = (1 + (,)— (20.4a)
Lt-1
We can rewrite this, dividing the numerator and denominator of the right side by E, and adding and subtracting I in thc denominator:
I + i'I = "г,+ ''''г (20.4b)
, A'-. - E>
We saw that this could be approximated by:
F1' - F
i, = i, - Z (20.4c)
4
Equivalently. the , + f( = (j +
which says that the domestic interest rate, it, must be equal to thc foreign interest rate, i*. minus thc Financial markets ► expected rate ol appreciation of the domestic currency, (£',', | - F.,)/E,. equilibrium. Condition 2 por nf)w we wj,| cake lhc cxpcttcc] future exchange rale as given and denote it as FA (Wc will relax (arbitragevThe expcctcd ^ assUmption in Chapter 21.) Under this assumption, and dropping time indexes, the exact interest parity condition (20.4a) becomes
rates of return on domestic and foreign bonds must be eqjal.
A
domestic interest rate ' ' E1
must equal the foreign
interest rate minus the Bringing £ to the left side on its own gives the current exchange rate as a function of the expected expected rate of future exchange rate, the domestic interest rate and the foreign interest rate:
appreciation of the
domestic currency. 0 + h1 - .
E, = — -jrE' (20,5)
(1 + h)
Equation (20.5) implies a positive relation between thc domestic interest rate and the current exchange rate, Civen the expected future exchange rate and the foreign interest rate, an increase in the domestic interest rate leads immediately to an increase in the exchange rate—equivalently, to an appreciation of the domestic currency. A decrease in the domestic interest rate leads immediately to a decrease in the exchange rate—to a depreciation of the domestic currency.
_> E". ► This relation between the exchange rate and thc domestic interest rate plays a central role in the i! => Ei. real world, and will play a central role in thc rest ot this chapter. To understand it further, think about the sequence of events that takes place in financial markets and foreign exchange markets after an increase in the Australian interest rate above the US interest rate:
• Start from a situation where the Australian and US interest rates are equal, so that i = ;*. This implies, trom equation (20.5), that the current exchange rate equals the expected future exchange rate: E = E'\
• Suppose, as a result ot an Australian monetary policy contraction, that the Australian interest rate increases. At an unchanged exchange rate, it becomes more attractive to hold Australian bonds, so financial investors want to shift out of US bonds and into Australian bonds. To do so, they must sell US bonds for US dollars, then sell US dollars for Australian dollars, then use thc Australian dollars to buy Australian bonds. As investors sell US dollars and buy Australian dollars, thc Australian dollar appreciates immediately.
• That an increase in the Australian interest rate leads to an appreciation of the Australian dollar is intuitively straightforward. An increase in the demand (or Australian dollars leads to an increase in the price of Australian dollars. What is less intuitive is by how much the Australian dollar must
appreciate. The important point here is: if financial investors don't change their expectation of the future exchange rate, then the more the Australian dollar appreciates today, the more investors expect it to depreciate in the future as they expect it to return to the same value in the future). Other things heing equal, this expectation makes LIS bonds more attractive. When the Australian dollar is expected to depreciate, a given rate ol return in US dollars means a higher rate of return in Australian dollars.
• This gives us the answer: the initial and immediate Australian dollar appreciation must be such lhat the expected future depreciation compensates for the increase in the Australian interest rate. When this is the case, investors are again indifferent and equilibrium prevails.
A numerical example will help. Assume lhat, until now, the one-year Australian interest rate and the one-year LIS interest rate were both equal to 4 per cent. Suppose that the Australian interest rate now increases to 10 per cent. II the expected exchange rate for next year doesn't change, the Australian dollar will appreciate by approximately 6 per cent today (or by exactly 5.46 percent). Why? Because, il the Australian dollar appreciates by approximately 6 per cent today and investors don't change their expectation of the exchange rate one year ahead, thc Australian dollar is now expected to depreciate by approximately 6 per cent over the coming year. Put the other way, the US dollar is expected to appreciate by approximately 6 per cent over thc Australian dollar during the coming year so that holding LIS bonds yields an expected rate of return of 10 per cent: the 4 per cent rate of return in LIS dollars, plus the expected approximate 6 per cent appreciation of the US dollar against the Australian dollar. Holding either Australian bonds or US bonds yields an expected rate of return of 10 per cent in Australian dollars, l inanciai investors are willing to hold either bond, so there is equilibrium in the foreign-exchange market.
E'n I ~ Ef
In terms of equation (20.4b),
(1 + 0
I + i, =
1 +
Et
+ 4%
+ 10% =
I - 5.46%
In terms of the approximate version, equation (20.4c),
F1' - F 4-1 4
4 Equation (20.5): E, = [(1 +!,)/< 1+!?)]£;.,. Rewriting with the interest rate on the left side gives: i, = (1 + /*,)£, /E!,»1 - 1. So. the slope of the line in Figure 20.1 is (1+ i*,)/EV, and the intercept is -I.What happens to the curve if,' increases.' If t'
тгголчрО
1, = I, -
E,
10% = 4% - (-6%)
The rate of return from holding Australian bonds (the left side) is equal to 10 per cent. Thc expected rate of return from holding US bonds expressed in Australian dollars (the right side) is approximately equal to thc US interest rate, 4 per cent, minus thc expected depreciation of thc Australian dollar, 6 per cent.
Figure 20.1 plots the relation between the (domestic > interest rate and the exchange rate implied by equation (20.5)—thc exact lorm of the interest parity relation. The relation is drawn for a given expected luture exchange rate, FJ, and a given foreign interest rate, Г. The lower the domestic interest rate, the lower the exchange rate. The relation is represented by an upward-sloping straight line, liquation (20.5) also implies that when the domestic interest rate is equal to the foreignjnteresl rate the exchange rate is equal to the expected future exchange rate: when i =_/*, then E = £''. This point is denoted as A in the figure. The slope of thc interest parity line is \ I - i*)/E1'. If cither thc foreign interest rate or the expected exchange rate change, the slope of the curve changes. (Note that it doesn't shift.)
Make sure you understand the steps in the argument:
• The one-year interest rate on Australian bonds increases by
6 per cent.
• Investors then buy Australian bonds.To pay for them, they must first buy Australian dollars.
• The Australian dollar appreciates until t
is expected to 4 depreciate by approximately 6 per cent during the coming year.
• This happens when the dollar has appreciated today by approximately 6 per cent
This example shows you 4 how approximate tie approximate' version of interest parity can be for even quite modest values of interest rates.
OUTPUT. THE INTEREST RATE ANDTHE EXCHANGE RATE
chapter 20
Note that our argument relies heavily on the assumption that, when the interest rate changes, the expected exchange rate remains unchanged. This implies that an appreciation today leads to an expected depreciation in thc future—as the exchange rate is expected to return to the same, unchanged.
Figure 20.1 The relation between the interest rate and the exchange rate implied by the exact interest parity relation
Interest parity relation (given i*, £')
«
£ о О
Ee Exchange rate, E
Depreciation of the domestic currency
Appreciation of the domestic currency
A tower domestic interest rate leads to a lower exchange rate—to a depreciabon of the domestic currency. A higher domestic interest rate leads to a higher exchange rale—to an appreciation of the domestic currency.
value in thc future. We will relax thc assumption that the future exchange rate is lixed in Chapter 21. But the basic conclusion will remain: An increase in the domestic interest rate relative lo the foreign interest rate leads to an immediate appreciation.
SUDDEN STOPS,THE STRONG US DOLLAR,THE VULNERABLE AUSTRALIAN DOLLAR AND THE LIMITS TO THE INTEREST
The interest parity condition assumes that financial investors care only about expected returns. As we discussed in Chapter 18, investors care not only about returns but also about risk and liquidity—how easy it is to buy or sell an asset.
Much of the time, we can ignore these other factors. Sometimes, however, these factors play a big role in investors' decisions and in determining exchange rate movements.
Perceptions of risk often play an important role in the decisions of large financial investors—-for example, pension funds—to invest or not to invest at all in a country. Sometimes the perception that risk has decreased leads many foreign investors to simultaneously buy assets in a country, leading to a large increase in demand for the assets of that country. Sometimes the perception that risk has increased leads the same investors to want to sell all the assets they have in that country, no matter what the interest rate.These episodes, which have affected many Latin American and Asian emerging economies, are known as sudden stops. During these episodes, the interest parity condition fails, and the exchange rate may decrease a lot, without any change in domestic or foreign interest rates.
Large countries can also be affected. For example, the appreciation of the US dollar in the 1990s—which, as we saw in Chapter 19, is one of the causes of the large US trade deficit—came not so much from an increase in US interest rates over foreign interest rates as from an increased foreign demand for dollar assets at a given interest rate. Many private foreign investors want to have some proportion of their wealth in US assets: they perceive US assets as being relatively safe. Many foreign central banks want to hold a large proportion of their reserves in US T-bills.The reason they do this is because theT-bill market is very liquid, so they can buy and sell T-bills without affecting the price. This very high demand for US assets, at a given
interest rate, was behind the 'strong dollar' in the 1990s. Even while US interest rates are relatively low. foreign investors are still eager to increase their holdings of US assets, and thus to finance the large US trade deficit. How long they are willing to do so will determine what happens to the US dollar and to the US trade balance.
Australia is certainly not immune to sudden stops. In October 2008 the Australian dollar fell almost 10 per cent in a day.This followed a surprisingly large interest rate cut of I per cent by the Reserve Bank of Australia (RBA) on 7 October. Such a large fall in the exchange rate cannot be accounted for by the interest parity condition. Figure I shows the daily movements over two months to October 2008 in the US$/A$ and the nominal effective exchange rate.The sudden stop was similar for both series, which suggests that this was a sudden loss of confidence in the Australian dollar generally. What could be the reason for this? In September and October 2008. world financial markets were in crisis, and the critical problem was the credit crunch— financial institutions everywhere had become very reluctant to lend. As you saw in Chapter 18, Australia has a very large and growing current account deficit (equal now to about 6 per cent of GDP). In 2008 this was equivalent to about A$200 million per day, which Australia needed to be able to borrow. When the RBA cut the cash rate by a huge I per cent in October 2008, fragile global lenders surely panicked, wondering if the RBA had advance information about problems in Australian financial institutions. On the day, foreign lenders stopped lending, and it took a dramatic fall in the currency before normal business was resumed.
OUTPUT. THl INTEREST RATF ANDTHL EXCHANGE RATF
chaptcr 20
The key lesson is that countries with a large net foreign debt and a high current account deficit are likely to be vulnerable to sudden stops in times of crisis.
0.85 -
0.80 -
0.75 -
US$/A$— left scale
Effective exchange rate right scale
■ 65
- 60
0.90 -i
Г 70
in 3
0.70
Figure I A sudden stop in the Australian dollar
0.65
■55
TT"
i i i i i i i i i i
l l l l l I l l ГТ
<
о — — —
The dashed line represents die day before the RRA cut the interest rate by one percentage point
SOURCE: Reserve Bank of Australia.
20.3 PUTTING GOODS AND FINANCIAL MARKETS TOGETHER
Financial market equilibrium is determined by thc equality ot money supply and money demand:
YL(i)
M P
Monetary policy is characterised by the simple choice of a particular value of thc interest rate:
i = «о
E, =
And the interest parity condition implies a positive relation between the domestic interest rale and the exchange rate:
(i + »,)
(I + i,)
Together, these lour relations determine output, the interest rate, the money supply and ihe exchange rale. Working with four relations isn't very easy. But we can easily reduce them to three by using thc interest parity condition to eliminate the exchange rate in the goods-market equilibrium relation. Doing this gives us ihe following three equations, the open-economy versions of our familiar IS-LM model:
. (1 + '»)- Y,Y -—г-£'• ( 1 + J, )
IS: Y = C(Y - Л + 1(Y, i) + C7 + NX
(20.6)
(20.7) 20.8)
M
LM: у = YLU)
Interest rate setting:
'n
Take thc IS relation lirst and consider the effects of an increase in the domestic interest rale on output. An increase in the interest rate now has two effects:
• The lirst ellect, which was already present in a closed economy, is the direct ellect on investment. A higher interest rate leads to a decrease in investment, so to a decrease in the demand lor domestic goods and a decrease in output.
• Thc second cllcci, which is only present in thc open economy, is thc effect through ihe exchange rate An increase in the domestic interest rale leads to an appreciation ol thc domestic currency. The appreciation, which makes domestic goods more expensive relative to foreign goods, leads to a dccrcasc in net exports, so to a dccrcasc in the demand tor domestic goods and a decrease in output. Both effects work in thc same direction. An increase in the interest rale decreases demand directly,
and indirectly, through thc adverse elfect ol ihe appreciation on demand. Therefore, the slope ol thc IS curve will he (latter in the open economy because ol the additional exchange rale ellcct.
directly and indirectly (through the exchange rate), to a decrease in output.
The IS relation between the interest rate and output is drawn in Figure 20.2. panel (a) lor given values of all the other variables in the relation, namely, T, G, Y*. i* and £' The IS curve is downward sloping an increase in the interest rate leads to a decrease in output. It looks very much the same as in thc closed economy, but it hides a more complex relation than before. The interest rate affects output An increase in the ► not only directly hut also indirectly through the exchange rate. interest rate leads, both
The LM relation is exactly the same as in the closed economy. The LM curve is upward sloping. For a given value ol the real money stock, MlP, an increase in output leads to an increase in the demand for money, and would require an increase in thc equilibrium interest rate. The central bank chooses thc interest rate to be fn, and so the money supply, AI. has to adjust endogenously so thai lor any given value ol output ihe money market is in equilibrium at that interest rale. The LM curve shilts up or down depending on whether the money supply has to be reduced or increased endogenously.
Equilibrium in the goods and linancial markets is attained at point A in panel 1 a), with output level Y and interest rate /„. I he equilibrium value ol the exchange rate cannot be read directly from the graph. Bui it is easily obtained Irom panel (hi, which replicates Figure 20.1 and gives the exchange
Figure 20.2 The IS-LM model in the open economy
(b)
(a)
LM
Interest parity relation
£ 'o
•- '0
E e Q
£ о О
Exchange rate, E
An increase in the interest rate reduces output both directly and indirectly (through the exchange rate): the IS curve is downward sloping. Given the real money stock, an increase in income requires an increase in the interest rate: the LM curve is upward sloping. Since the central bank chooses the interest rate, the money supply is endogenous. If the domestic interest rate rises, the exchange rate appreciates: the interest parity relation is upward sloping.
rate associated with a given interest rate. The exchange rate associated with thc interest rate to is equal to E.
To summarise: We have derived the IS and thc LM relations lor an open economy. Thc IS curve is downward sloping: An increase in the interest rate leads directly and indirectly through the exchange rate to a decrease in demand and a decrease in output.
The LM curve is upward sloping-. An increase in income increases the demand lor money, requiring an increase in the equilibrium interest rate.
With the central bank choosing the interest rate, i(), equilibrium output is read ofl the IS curve, and thc quantity ot money, M, adjusts endogenously so that the LM curve goes through thc point ■ x0,V). Given the foreign interest rate and the expected future exchange rate, the domestic interest rate determines the equilibrium exchange rale front the interest parity relation.
20.4 THE SHORT-RUN EFFECTS OF POLICY IN AN OPEN ECONOMY
A monetary policy contraction leads to a 4 higher interest rate, requiring the money stock to be reduced, thus shifting the LM curve up. It shifts neither the IS curve nor the interest-parity curve.
Having derived the IS-LM model lor the open economy, we now put it to use and look at the effects of policy. First, wc will consider a change to monetary policy. Then we will change liscal policy, considering the situation when thc central bank docs or docs not respond to the effects of the fiscal policy.
The short-run effects of monetary policy in an open economy
Let's look at the ellects ol a monetary policy contraction (with a constant fiscal policyI Look at Figure 20.3, panel (a). At a given level ol output, an increase in the interest rate to i" requires open- market operations by thc central bank to shift the LM curve up from LM to LM'. The equilibrium moves Irom point A to point A'. In panel 1 h), the increase in thc interest rate leads to an appreciation ot the domestic currency.
Output, V
So, a monetary policy contraction leads to an increase in the interest rate, to a decrease in output, ami to an appreciation of the domestic currency. The story is easy to tell. A monetary contraction leads to an increase in the interest rate, making domestic bonds more attractive and triggering an appreciation. The higher interest rate and the appreciation both decrease demand and output. Investment lalls because of thc higher interest rate and lower output and consumption falls because of lower output.
Figure 20.3 The effects of a monetary policy contraction
0 л
ы
* ,
£ i
1
U
3
(/) 0)
У' У
Output, У
(b)
Interest parity relation .
£ £' Exchange rate, E
Appreciation ►
£ о О
A monetay policy contraction /eods to a decrease in output and on appreciation.
Thc effect on imports (ancf thus the trade account) is ambiguous because the output lall reduces imports but thc appreciation raises imports (and reduces exports, i
The short-run effects of fiscal policy in an open economy
Now let's look at our other favourite policy experiment, a change in government spending. Suppose that, starting from a balanced budget, the government decides to increase delence spending without raising taxes, and so ains a budget deficit. What happens to the level ol output? To the interest rate? To the composition ot output? To thc exchange rate?
Ihe answers arc given in Figure 20.4. The cconomy is initially at point A Thc increase in government spending by AG > 0 increases output at a given interest rate shifting the IS curve to thc An increase in ► right, from /5 to IS' in panel (a). Because government spending doesn't enter thc LM relation, thc LM government spencing curve doesn't shift. What happens to the interest rate? As we saw in Chapter 5, it depends on the cads to an increase in response of the central bank lo ibis aggregate demand shock:
The central bank decides not to respond. The interest rate will remain unchanged at i - /',,. and thc outcome is identical lo what happened in our closed cconomy analysis. The new equilibrium is at point A', with a higher level ol output, V. The exchange rale doesn't change. The increase in output raises consumption, investment and imports, and as a consequence net exports worsen. Thus, thc fiscal delicit leads to a trade delicit.
Thc ccntral bank recogniscs that the higher output will cause inflationary pressure in the economy and so decides to raise the interest rate lo i". In panel (b) we show i" so that the new equilibrium A" is on the original LM curve. We did this simply to keep thc number ol curves in ihe diagram lo a minimum—remember, the central hank is Iree to choose anv value tor i". and the new LM curve would have to adjust accordingly. The higher interest rate leads to an increase in thc exchange rate—an appreciation ol ihe domestic currency. So, an increase in government spending leads to an increase in the interest rate, an increase in output tless than when the central hank didn't respond/, and an appreciation of the domestic currency.
output, and an appreciation rf the central bank responds with a higher interest rate. An increase in government spending shifts the IS curve to the t
right A central bank response requires a shift up of the LM curve.The interest-parity curve doesn't shift (since we have assumed the expectcd future exchange rate is fixed).
In words: An increase in government spending leads to an increase in demand, leading to an increase in output. /Is output increases, the central bank is prompted to raise the interest rate. The increase in the interest rate, which makes domestic bonds more attractive, leads to an appreciation of the domestic currency. The higher interest rate and the appreciation of the domestic currency both
An increase in government spending leads to an increase in output and, if the central bank responds, to a higher interest rate and an appreciation.
decrease the domestic demand for goods, offsetting some of the effect of government spending on demand and output.
Can we tell what happens to the various components ot demand?
• Clearly consumption and government spending both go up—consumption goes up because of the increase in income.- government spending goes up by assumption.
• What happens to investment is ambiguous. Recall that investment depends on both output and the interest rate: I = IlY.i . On thc one hand, output goes up. leading to an increase in investment. But on the other the interest rate also goes up. leading to a decrease in investment. Depending on which ol these two effects dominates, investment can go up or down. In short: when the central bank responds to the effects of tiscal policy, thc effect of government spending on investment was ambiguous in the closed economy,- it remains ambiguous in the open economy.
• Recall that net exports depend on domestic output, foreign output and the exchange rate: NX - NXi Y,Y*,L). Thus, the appreciation and the increase in output combine to decrease net exports: the appreciation decreases exports and increases imports, and the increase in output further increases imports. So. the budget deficit leads to a deterioration ol the trade balance. Il trade is balanced to start with, then the budget delicit leads to a trade deficit. Note that, while an increase in the budget dclicit increases thc trade delicit, the effect is lar from mechanical. It works through the ellect ol the budget delicit on output anil on the exchange rate and, in turn, on the trade dclicit.
This modernised version ol the IS-I.X 1 model lor the open economy is based on a similar model first put together in thc 1960s by two economists, Robert Mundell, at Columbia University, and Marcus Fleming, at the IMF. For this reason, it is called the .Wundell-I'leming model. How well does it lit the 4 Robert Mundell received tacts? To answer, one could hardly design a better experiment than the sharp monetary and fiscal policy the Nobel Prize for changes that the US economy underwent in the early 1980s. The evidence is shown in the focus box economics in 1999. 'Monetary contraction and fiscal expansion: Thc United Stales in the early 1980s. Conclusion: the Mundell Fleming model and its predictions pass with (lying colours.
У V" У' Output, У
£ E" Exchange rate, E
Figure 20.4 The effects of an increase in government spending
Lets see il some ol thc implications ot the Mundell-Fleming model are useful lor understanding what has happened in Australia in thc last twenty years. We discussed various policy mixes in the Chapter 5 foctis box The Australian policy mixes, 1986-2008'. We leaturcd four episodes—1986-91, 1992-95, 1996—2001 and 2002-08. The lirst mix was characterised by tight liscal and tight monetary
policy, the second hy easy fiscal and easy monetary policy, the third by tight fiscal and easy monetary policy, and in the fourth both were tight until the global linancial crisis became acute in October 2008. Regardless ol thc particular mix, thc Mundcll—Fleming model suggests two important relationships:
• A positive relationship between fiscal and trade deficits: This is known as thc twin deficits hypothesis A liscal expansion—that is, an increase in the fiscal deficit—raises output and thus imports, which creates a trade delicit. If thc central bank raises interest rates in response to the higher output, thc exchange rate appreciates, which itself raises imports and lowers exports—the net effect of higher output and the exchange rale is always a worse trade account. Thc immediate evidence lor these two relationships isn't strong. From our discussion in Scction 19.6. you shouldn't expect a strong link. First, it is possible that the trade balance is a target ol fiscal policy. Il the trade balance worsens, the government may choose to raise the fiscal surplus. Also, it is quite possible that private saving and investment may counteract and sometimes even dominate the ellect ol the liscal balance on the trade balance. Only in the unlikely circumstance that domestic saving always lully finances investment would the link be perfect. And finally, global commodity prices are exogenous to Australia and have a substantial impact on the trade account.
• A positive relationship between the real interest rate and the real exchange rate: Whenever the central bank raises interest rates, the exchange rate will appreciate.
FOCUS f Д
rBOX
MONETARY CONTRACTION AND FISCAL EXPANSION: HE UNITED S
Thc cvidcnce for this relationship between thc real interest rate and thc real exchange rate is typically stronger.
Д 'ДД1 v*': '*:** a л
The early 1980s in the United States were dominated by sharp changes in both monetary policy and fiscal policy.
By the late 1970s, the chairman of the Fed, Paul Volcker, concluded that US inflation was too high and had to be reduced. Starting in late 1979.Volcker embarked on a path of sharp monetary contraction. He realised that this might lead to a recession in the short run. but predicted it would lead to lower inflation in the medium run.
The change in fiscal policy was triggered by the election of Ronald Reagan in 1980. Reagan was elected on the promise of more conservative policies—namely, a scaling down of taxation and the government's role in economic activity. This commitment was the inspiration for the Economic Recovery Act of August 1981. Personal income taxes were cut by a total of 23 per cent, in three instalments from 1981 to 1983. Corporate taxes were also reduced. These tax cuts were not, however, accompanied by corresponding decreases in government spending, and the result was a steady increase in budget deficits, which reached a peak in 1983 at 6.0 per cent of GDP. Table I gives spending and revenue numbers for 1980-84.
What were the Reagan administration's motivations for cutting taxes without implementing corresponding cuts in spending? These are still being debated today, but there is agreement that there were two main motivations.
One motivation came from the beliefs of a fringe, but influential, group of economists called the supply- siders. who argued that a cut in tax rates would lead people and firms to work much harder and more productively, and that the resulting increase in activity would lead to an increase, not a decrease, in tax revenues. Whatever the merits of the argument appeared to be then, it proved wrong. Even if some people did work harder and more productively after the tax cuts, tax revenues decreased and the fiscal deficit increased.
The other motivation was the hope that the cut in taxes, and the resulting increase in deficits, would scare the US Congress into cutting spending or, at the very least, into not increasing spending further. This motivation turned out to be partly right: Congress found itself under enormous pressure not to increase spending, and the growth of spending in the 1980s was surely lower than it would have been otherwise. Nonetheless, this decrease in spending wasn't enough to offset the shortfall in taxes and avoid the rapid increase in deficits.
Table I The emergence of large US budget deficits. 1980-84
1980 1981 1982 1983 1984
Spending 21.7 22.2 23.1 23.5 22.1
Revenues 19.0 19.6 19.2 17.4 17.3
Personal taxes 9.0 9.3 9.2 8.4 7.8
Corporate taxes 2.4 2.0 1.5 I.I 1.5
Budget surplus (- = deficit) -2.7 -2.8 ^1.0 -6.0 -4.8
Numbers are for fiscal years, which start in October of the previous calendar year. All numbers are expressed as a percentage of G DP.
SOURCE: Historical Tables. Office of Management and Budget.
Whatever the reasor for the deficits, the combined effects of the monetary contraction and the fiscal expansion were in line with what the Mundell-Fleming model predicts.Table 2 gives the evolution of the main macroeconomic variables from 1980 to 1984.
From 1980 to 1982, the evolution of the economy was dominated by the effects of the monetary contraction. Interest rates, both nominal and real, increased sharply, leading to a large US dollar appreciation (an increase in the exchange rate) and to a recession.The goal of lowering inflation was achieved, although not right away: by 1982, inflation was down to about 4 per cent. Lower output and dollar appreciation had opposing effects on the trade balance (lower output leading to lower imports and an improvement in the trade balance: the appreciation of the dollar leading to a deterioration in the trade balance), resulting in little change in the trade deficit before 1982.
From 1982 on, the evolution of the economy was dominated by the effects of the fiscal expansion. As our model predicts, these effects were high interest rates, strong output growth and further dollar appreciation. The effects of high output growth and dollar appreciation were an increase in the trade deficit to 2.7 per cent of GDP by 1984. By the mid-1980s. the main macroeconomic policy issue had become that of the twin deficits, the budget deficit and the trade deficit. It was to remain one of the central macroeconomic issues throughout the 1980s and in the first part of the 1990s.
Table 2 Major US macroeconomic variables. I980--84
1980 1981 1982 1983 1984
GDP growth (%) -0.5 1.8 -2.2 3.9 6.2
Unemployment rate (%) 7.1 7.6 9.7 9.6 7.5
Inflation (CPI) (%) 12.5 8.9 3.8 3.8 3.9
Interest rate (nominal) (%) 11.5 14.0 10.6 8.6 9.6
(real) (%) 2.5 4.9 6.0 5.1 5.9
Real exchange rate 85 101 112 118 130
Trade surplus -0.5 -0.4 -0.6 -1.5 -2.7
(- = deficit) (% of GDP)
1 1
Inflation: Rote of change of trie CPI. The nominal interest rate is the three-month T-bill rate. The real interest rate is equal to the nominal rate minus the forecast of inflation by DRI, a private forecasting firm. The reol exchange rate is the US multilateral real exchange rate, normalised so that 1973 = 100.
20.5 FIXED EXCHANGE RATES
Wc have assumed so lar that the central bank chose the money supply and let the exchange rate adjust in whatever manner was implied by equilibrium in thc foreign-exchange market. In most countries, this assumption doesn't reflect reality: central banks aci under implicit or explicit exchange rate targets ard use monetary policy to achieve those targets. The targets are sometimes implicit, sometimes explicit,- thev arc sometimes specific values, sometimes bands or ranges. These exchange rate arrangements 'or regimes as ihey arc callcd) come under many names. Let's first see what these names mean.
Pegs, crawling pegs, bands, the EMS and the euro
At one end ol thc spectrum arc countries with flexible exchange rates such as the LIniicd Stales, Australia and Japan. These countries have no explicit exchange rate targels. While their central banks don't ignore movements in thc exchange rale they have shown themselves quite willing to let their exchange rales fluctuate considerably.
At the other end are countries that operate under fixed exchange rates. These countries maintain a fixed exchange rate in terms ol some foreign currency. Some peg their currency to the dollar, l or example, from 1941 to 2001 Argentina pegged its currency, the peso at the highly symbolic exchange rate of one LIS dollar lor one peso more on this in Chapter 21 . and in our region, the 1 long Kong dollar has been pegged to the LIS dollar at HK$7.8 per US$1. Other countries used to peg their currency lo the French franc (most ol these are former French colonics in Africaas the French franc has been replaced by the euro, they arc now pegged to the euro. Yet others peg to a basket of currencies, with thc weights reflecting the composition ol their trade.
Thc label 'fixed' is a bit misleading. It isn't the case thai the exchange rate in countries with fixed exchange rates actually never changes. But changes are rare. An extreme case is that ol the African countries pegged to thc French franc. When their exchange rates were readjusted in January 1994, this was the lirst adjustment in forty-five years. Because these changes are rare, economists use specific words to distinguish them from thc daily changes that occur under flexible exchange rates. They rcler to a dccrcasc in the exchange rate under a regime of fixed exchange rates as a devaluation rather than a depreciation, and to an increase in the exchange rate under a regime of fixed exchange rates as a revaluation rather than an appreciation.
Between these extremes arc countries with various degrees of commitment to an exchange rate target. For example, some countries operate under a crawling peg. Thc name describes it well. I hese countries often have inllation rales that exceed ihe LIS inflation rate. II they were to peg their nominal exchange rale against the LIS dollar the increase in their domestic pricc level relative to the LIS price level would lead to a steady real appreciation and rapidly make their goods uncompetitive with LIS goods. To avoid this elfect, these countries choose a predetermined rate ol depreciation against the LIS dollar. They choose to 'crawl imovc slowly) with relation lo thc dollar.
Yet another arrangement is lor a group ol countries to maintain their bilateral exchange rates I thc exchange rate between each pairoi countries) within some bands. Perhaps the most prominent example- was the European Monetary System (EMS), which determined thc movements ol exchange rales within thc European Llnion trom 1978 to 1998. Llndcr I MS rules, member countries agreed to maintain their exchange rate in relation to the other currencies in the system within narrow bands around a central parity—a given value for the exchange rate. Changes in thc central parity and devaluations or revaluations ol specilic currencies could occur, bur only by common agreement among member countries. After a major crisis in 1992, which led several countries to drop out ol the EMS altogether, exchange rate adjustments became more and more infrequent, leading several countries to move one- step further and adopt a common currency, thc euro. Conversion Irom domestic currencies to the euro started on I January 1999 and was completed in early 2002 by fifteen European countries. Wc will return to the implications ol the move to the euro in Chaptcr 21.
Like the 'dance of the US dollar' in the 1980s (Chapter 18). there was a 'dance of the Australian dollar' and a 'dance of the yen' in the 1990s. From Figure 20.6. we can see how the Australian dollar first depreciated, then appreciated and then depreciated during the 1990s. There was a sharp appreciation of the yen in the first half of the 1990s. followed by a sharp depreciation later in the decade.
Recall the definition of the real exchange rate, e = EPIP*. If domestic inflation is higher than foreign inflation, then:
• P increases faster
than P*.
• Equivalendy. PIP'
increases.
• Iff is fixed. EPIP*
increases. ► Equivalently. there is a steady real appreciation: domestic goods become steadily more expensive relative to foreign goods.
We look at the 1992 ► crisis in Chapter 21.
You can think of countries adopting a common currency as ^ adopting an extreme form of fixed exchange rates: their exchange rate' is fixed at one to one between any pair of countries.
We will discuss the pros and cons of these dilferent exchange regimes in the next chapter. Bui first you must understand how pegging ihe exchange rate allects monetary policy and fiscal policy. This is what we do in the rest ol this section.
Pegging the exchange rate, and monetary control
Suppose a country decides :o peg its exchange rate at some chosen value—call it £. How does it actually achieve this? Thc government cannot just announce the value ol the exchange rate and stand there. Rather, it must take measures so that its chosen exchange rate will prevail in thc foreign-exchange market. Lets look at this more closely.
JL.
Pegging or no pegging, the exchange rate and the nominal interest rate must satisfy the interest parity condition:
I + i, - (I
Now suppose that the country pegs the exchange rate at £, so the current exchange rate C, = £. If linanciai and foreign-exchange markets believe that the exchange rate will remain pegged at this value, then their expectation of the luture exchange rate, E',. . is also equal to£, and the interest parity relation becomes
I
I
/, = (i + 0=
In words: If financial investors expect the exchange rate to remain unchanged, the}/ will require the same nominal interest rate in both countries. Llnder a fixed exchange rate and perfect capital mobility, the domestic interest rate must be equal to lhe foreign interest rate.
4 Under perfect capital mobility, fixing the exchange rate means giving up the freedom to choose the domestic interest rate, which must remain equal to the foreign interest rate.
_M P
(20.9)
These results depend very much on the interest rate parity condition, which in turn depends on the assumption of perfect capital mobility. (Financial investors go for the highest expected rate of return.) The case ^ of fixed exchange rates with imperfect capital mobility—which is more relevant for middle- income countries, such as in Latin America or Asia—is treated in Appendix 2 to this chapter.
This condition has one further important implication. Return to the equilibrium condition lhat the supply of money and demand for money be equal. Now that i - /*, this condition becomes
YL(i*)
Suppose an increase in domestic output increases the demand (or money. In a closed economy, thc central bank could leave the money stock unchanged, leading to an increase in thc equilibrium interest rate, or it could choose its desired interest rate and adjust the money stock to achieve equilibrium. In an open economy, and under flexible exchange rates, we have seen that the central bank still has the same choices. We assumed, in line with contemporary practice, that thc central hank chooses the interest rate based on an inflation target rule. Thus, an increase in output will result in an increase in the interest rate and an appreciation ot the domestic currency. But under fixed exchange rates, the central bank cannot choose the interest rate independent of the exchange rate and the loreign interest rate, nor can it keep the n-.onev stock unchanged. It it did, the domestic interest rale would increase above the foreign interest rate, leading to an appreciation ot thc domestic currency. To maintain thc exchange rate, the central bank must increase the supply of money in line with the increase in the demand tor money so the equilibrium interest rale doesn't change. Given the price level, P, nominal money, AI. must adjust so that equation (20.6) holds.
To summarise: llnder fixed exchange rates, the central bank gives up monetary policy as a policy instrument. A lixed exchange rate implies a domestic interest rale equal to the foreign rate. And the money supply must adjust to maintain the interest rale. Or, we can say lhat the central bank's monetary policy targets ihe exchange rale that it hopes ю achieve at all times. It can succeed as long as it owns sufficient foreign exchange reserves. The only possible monetary policy change in a fixed exchange rate- regime is a change in thc chosen exchange rate. We discuss this in Chapter 21.
Fiscal policy under fixed exchange rates
It monetary policy can no longer be used under lixed exchange rales, what about tiscal policy? To answer this, wc use Figure 20.5.
Figure 20.5 starts by replicating Figure 20.4. panel ia). which wc used earlier to analyse thc effects ol liscal policy under flexible exchange rates. In that case we saw that a liscal expansion AC > 0)
shifted the IS curve to the right. Under flexible exchange rates with the central bank raising the interest rale lo counter the inflationary consequences of higher demand, this leads to a movement in the equilibrium from point А lo point П. with an increase in output Irom Ул to Y(i, and an increase in the exchange rate—an appreciation ol the domestic currency.
However, under fixed exchange rates ihe central bank cannot lei the currency appreciate. As the increase in output leads it) an increase in ihe demand for money, thc central bank must accommodate this increased demand tor money by increasing the money supply. In terms of Figure 20.5 the ccntral bank must shilt the LM curvc down as ihe IS curve shifts to thc right so that the interest rate, and thus the exchange rale, don't change. The equilibrium therefore moves trom Л to C, with higher output Y( and unchanged interest and exchange rates. So under fixed exchange rates, fiscal policy is more powerful than il is under flexible exchange rates. This is because liscal policy triggers a monetary policy response under flexible exchange rates, but monetary accommodation under fixed rates.
As this chaptcr comes to an end, a question should have started to form in your mind. Why would a country choose to fix its exchange rater You have seen several reasons why this appears to be a bad idea:
• By fixing thc exchange rate, a cotinirv gives up a powerful tool for correcting trade imbalances or charging ihe level of economic activity.
• By committing to a particular exchange rale, a country also gives up control ot its interest rate. Not only thai, but thc country must match movements in the loreign interest rate at the risk of unwanted effects on its own activity. This is what happened in the early 1990s in Europe. Because of the increase in demand due lo reunification, Germany felt il had to increase its interest rate. To maintain their parity with the deutschmark 1 DM 1 the German currency at the time . other countries in the European Monetary System were also lorccd to increase their interest rale something that they would rather have avoided. This is ihe topic ot ihe focus box German unification, interest rates and ihe EMS'.)
Figure 20.5 The effects of a fiscal expansion under fixed exchange rates
Under fixed exchange rotes, output increases from YA to Yc. Under flexible exchange rotes, a fiscal expansion increases output by less if the central bank chooses to increase the interest rate—for example to B.
This result—that the central bank needs to raise the interest rate, thus appreciating the currency and moderating the output increase—depends on our simplifying ► assumption that the future expected exchange rate is constant. In the next chapter we relax that assumption and it becomes possible that fiscal policy has no effect on output under flexible exchange rates, ever in the short run.
Is the effect of fiscal ► policy stronger in a closed economy or in an open economy with fixed exchange rates?
• While thc country retains control of fiscal policy, one policy instalment isn't enough. As you saw in Chapter 19, for example, a fiscal expansion can help the economy gel out ot a recession, but only
ai the cost of a larger trade dclicit. And, under lixed exchange rates, a country that wants, lor example, to decrease its budget delicit cannot use monetary policy to offset the contractionary effect of its fiscal policy on output.
FOCUS [BOX
So, why do some countries fix their exchange rate? Why have sixteen European countries recently adopted a common currency? To answer these questions, we must do some more work. Wc must look at what happens not only in the short run - -which is what wc did in this chapter—hut also in thc medium run, when the price level can adjust. We must look at the nature ol exchange rate crises. Once we have done all this, wc will then he able lo give an assessment of the pros and cons of exchange rate regimes. These are thc topics we take up in Chapter 21.
GERMAN UNIFICATION, INTEREST RATES AND THE EMS
Under a fixed exchange rate regime such as the European Monetary System (EMS) (let's ignore here the degree of flexibility that was afforded by the bands), no individual country can change its interest rate if the other countries don't change theirs as well. So, how do interest rates actually change? Two arrangements are possible: one is for all the member countries to coordinate changes in their interest rates; another is for one of the countries to take the lead and for the other countries to follow—this is what happened in the EMS, with Germany as the leader.
During the 1980s, most European central banks shared similar goals and were happy to let the Bundesbank (the German central bank) take the lead. But in 1990, German unification led to a sharp divergence in goals between the Bundesbank and the other EMS nations' central banks. The need for large transfers to eastern Germany and an investment boom led to a large increase in demand in Germany. The Bundesbank's fear that this shift would generate too strong an increase in activity led it to adopt a restrictive monetary policy. The result was strong growth in Germany together with a large increase in interest rates.
This may have been the right policy mix for Germany. But for other countries, this policy mix was much less appealing. The other countries hadn't experienced the same increase in demand, but to stay in the EMS they had to match the high German interest rates. The net result was a sharp decrease in demand and in output in the other countries.These results are presented in Table I, which gives nominal interest rates, real interest rates, inflation rates and GDP growth from 1990 to 1992 for Germany and for two of its EMS partners. France and Belgium.
Note first how the high German nominal interest rates were matched by both France and Belgium. Nominal interest rates were actually higher in France than in Germany in all three years! This is because France needed higher interest rates than Germany to maintain the DM/franc parity; the reason is that financial markets weren't sure that France would actually keep the parity of the franc against the DM. Worried about a possible devaluation of the franc, financial investors asked for a higher interest rate on French bonds than on German bonds.
While France and Belgium had to match—or, as we have just seen, more than match—German nominal rates, both countries had less inflation than Germany.The result was very high real interest rates, higher than in Germany. In both France and Belgium, average real interest rates from 1990 to 1992 were close to 7 per cent. And in both countries, the period 1990-92 was characterised by slow growth and rising unemployment. Unemployment in France in 1992 was 10.4 per cent, up from 8.9 per cent in 1990. The corresponding numbers for Belgium were 12.1 per cent and 8.7 per cent.
While we have looked at only two of Germany's EMS partners, a similar story was unfolding in the other EMS countries. By 1992. average unemployment in the European Union, which had been 8.7 per cent in 1990, had increased to 10.3 per cent.The effects of high real interest rates on spending weren't the only source of this slowdown, but they were the main one.
By 1992, an increasing number of countries were wondering whether to keep defending their EMS parity or to give it up and lower their interest rates. Worried about the risk of devaluations,financial markets started to ask for higher interest rates in those countries where they thought devaluation was more likely.The result was two major exchange rate crises, one in late 1992 and the other in mid-1993. By the end of these two
crises, two countries. Italy and the United Kingdom, had left the EMS.We will look at these crises, their origins and their implications, in Chapter 21.
Table 1
1990-92 German unification, interest rates and output growth; Germany, France and Belgium,
Nominal interest rate (%) 1990 1991 1992 1990 Inflation (%) 1991 1992
Germany 8.5 9.2 9.5 2.7 3.7 4.7
France 10.3 9.6 10.3 2.9 3.0 2.4
Belgium 9.6 9.4 9.4 2.9 2.7 2.4
1990 ■teal interest rate (% 1991 1992 1990 GDP growth (%) 1991 1992
Germany 5.7 5.5 4.8 5.7 4.5 2.1
France 7.4 6.6 7.9 2.5 0.7 1.4
Belgium 6.7 6.7 7.0 3.3 2.1 0.8
1
The nominal interest rate is the short-term nominal interest rate. The real interest rate is the realised real interest rate over the year-that is. the nominal interest rate minus actual inflation over the year. All rates are annual.
SOURCE OECD Economic Outlook.
SUMMARY
• In an open economy, the demand for goods depends on both the interest rale and the exchange rate. A decrease in the interest rale increases the demand for goods. An increase in ihe exchange rate— an appreciation—decreases the demand for goods.
• Financial market equilibrium is given by the equality of money demand and money supply. Under flexible exchange rates, the central bank chooses the inierest rale based on an inflation targeting rule. Thc exchange rate is determined by the interest parity condition, which stales lhal thc domestic interest rate must equal the foreign interest rate minus thc cxpectcd rate ol appreciation.
• Given the expected future exchange rate and thc foreign interest rate, increases in the domestic interest rate lead to an increase in the exchange rale (an appreciation). Decreases in the domestic interest rate lead to a decrease in the exchange rate la depreciation).
• Uncer flexible exchange rates an expansionary fiscal policy leads lo an increase in output, lo an increase in thc interest rate assuming the central bank responds to the implied inflationary pressure, and thus to an appreciation. A contractionary monetary policy leads 10 an increase in the interest rate, to an appreciation and thus to a decrease in output.
• There are many types of exchange rate arrangements. They range Irom fully flexible exchange rates to crawling pegs, pegs, fixed exchange rales and the adoption ol a common currency. Under fixed exchange rates, a country maintains a fixed exchange rate in terms ot a foreign currency or a basket of currencies.
• Under lixed exchange rates and thc interest parity condition, a country must maintain an interest rale equal lo the loreign interest rate. Thc central bank loses thc use of monetary policy as a policy instalment. However, fiscal policy becomes more powerful than under flexible exchange rates,
because liscal policy triggers monetary accommodation and so doesn t lead to offsetting changes in the domestic interest rate and exchange rate.
KEYTERMS
• Mundell-Fleming model, 455
• flexible exchange rates, 457
• perfect capital substitutability. 457
• perlcct capital mobility, 457
• sudden stops. 460
• twin deficits hypothesis 466
• supplv-siders. 466
• peg, 468
• crawling peg, 468
• European Monetary System (EMS), 468
• bands, 468
• central parity. 468
• euro, 468
QUESTIONS AND PROBLEMS
Quick check
1. Using the information in this chapter, label each of lhe following statements 'true', 'false' or 'uncertain'. Explain briefly.
a. A liscal expansion tends to an increase in net exports.
b. Fiscal policy has a greater effect on output in an economy with fixed exchange rates than in an economy with flexible exchange rates.
c. Other things being equal, the interest parity condition implies lhat the domestic currency will appreciate in response to an increase in the expected exchange rate.
d. Il linanciai investors expect the Australian dollar to depreciate in relation to the yen over the coming year, one-year interest rates will be higher in Australia than in Japan.
e. Il the Japanese interest rate is equal to zero, foreigners won't want lo hold Japanese bonds. I. Llnder tixed exchange rates, the money stock must be constant.
2. Consider an open economy with flexible exchange rates. Suppose that output is at the natural level, but there is a trade deficit.
What is the appropriate fiscal policy-monetary policy mix?
3. In this chapter we showed that a monetary expansion in an economy operating under flexible exchange rates leads to an increase in output and a depreciation of the domestic currency.
a. How does a monetary expansion (in an economy with flexible exchange rates1 affect consump¬tion and investment?
b. Flow docs a monetary expansion in an economy with flexible exchange rates I affect net exports?
4. Flexible exchange rates and foreign macroeconomic policy
Consider an open economy with flexible exchange rates. Let IP stand for the (uncovered) interest parity condition.
a. In an IS-LM—IP diagram, show the effect of a decrease in foreign output, Y*, on domestic output, Y, Explain in words.
b. In an IS-LM-1P diagram, show the ellect of an increase in the foreign interest rate, i*, on domestic output, Y. Explain in words.
c. Given the discussion ol the effects ol liscal policy in this chapter, what elfect is a foreign liscal expansion likely to have on foreign output Y" and on the foreign interest rate, /*? Given the discussion of the ellects ol monetary policy in this chapter, what effect is a foreign monetary expansion likely to have on Y* and /*?
d. Given your answers to pans (a). (b) and (c), bow does a foreign liscal expansion allcct domestic output? How docs a foreign monetary expansion allcct domestic output? (Hint: One ol these policies has an ambiguous effect on output.
Dig deeper
5. Fixed exchange rales and foreign macroeconomic policy
Consider a fixed exchange rate system, in which a group of countries hailed follower countries/ peg their currencies to the currency of one country (called the leader country). Since the currency of the leader country is not fixed against the currencies of countries outside the fixed exchange rate system, the leader country can conduct monetary policy as il wishes. For this problem, consider the domestic country to be a follower country and the foreign country to be the leader country.
a. Redo ihe analysis of problem 4(a).
b. Redo the analysis of problem 4(b >.
c. Using your answers to parts la) and :b and problem 4(c), how does a foreign monetary expansion (by the leader country) affect domestic output? How docs a foreign liscal expansion (by the leader country) alfcct domestic output? (You may assume that the etlcct ol Y* on domestic output is small How do your answers differ from those in 4(d)?
6. The exchange rate as an automatic stabiliser
Consider an economy that suffers a fall in business confidence (which tends to lower investment).
a. Suppose that the economy has a flexible exchange rate. In an IS-LM-IP diagram, show the short-run effect ot the tall in business confidence on output, the interest rate and the exchange rate. I low does ihe change in the exchange rate, by itsell, tend to affect output? Does the change in the exchange rale dampen (make smallerl or amplily < make larger the ellect ot ihe lall in business confidence on output?
b. Suppose instead thai the economy has a lixed exchange rate. In an IS-LM-IP diagram, show how the economy responds to the fall in business conlidencc. What must happen to thc money supply in order to maintain the lixed exchange rate? How does thc effect on output in this economy with fixed exchange rates compare with the effect you found for the economy in pari (a), which had llexible exchange rates?
c. F.xplain how the exchange rate acts as an automatic stabiliser in an economy with flexible exchange raics.
Explore further
7. In 2008-09 the Australian economy was affected by the global financial crisis. In Appendix I of this chapter, we introduce a relative risk premium, r, into the interest parity relation. Suppose there was a big rise in Australia's r in late 2008.
a. What docs thc IS-LM-IP model predict should be ihe ellecls ol this shock to ihe Australian economy?
b. The Australian government responded with a policy mix of expansionary monetary and liscal policy. Show how these policy responses allcct the equilibrium exchange rate and output in the model.
8. Expected depreciation of the LIS dollar
In Chapter 19 the text mentions that the LIS dollar may need to depreciate by as much as 20-40 per cent in real terms to achieve a reasonable improvement in the trade balance.
a. Go the website of The Economist (www.economist.com) and find the latest data on ten-year interest rates. I.ook in the section 'Markets & Data and then the subsection 'Economic and Financial Indicators', and finally Trade, exchange rates, budget balances and interest rates'. Look at thc interest rales for thc United States, lapan, China, Britain, Canada Australia and ihe euro area. For each country treating the euro area as a country), calculate the spreads as that country's interest rale minus the US interest rate, b. From the uncovered interest parity condition, the spreads from part a) are the annualiscd expected appreciation rates ol the dollar against other currencies. To calculate the ten-year expected appreciation, you must compound. (So, il x is the spread, the ten-year expected appreciation is |il л)"' I . lie carelul about decimal points. Is the dollar expected to depreciate much in nominal terms against any currency other than the yen? с (iiven your answer to part h il we accept that signilicant real depreciation ol the dollar is likely
in the next decade, how must it he accomplished? Does your answer seem plausible? d. What do your answers to parts h and с suggest about the relative strength ol demand for US dollar assets independent ol the exchange rate? You may want lo review Appendix I belore answering this question, interpreting a low relative risk premium perhaps a discount p. as a measure ot the relative strength of demand for US dollar assets.
We invite you to visit the Blanchard-Sheen page on the Pearson Australia website at
www.pearson.com.au/highered/blanchardsheen3e
for many World Wide Web exercises relating to issues similar to those in this chapter.
FURTHER READINGS
• A lascinating account ol the politics behind liscal policy under the Reagan administration in the United States is given by David Stockman—who was then the director ol the Ollice ol Manage¬ment and Budget OMB)—in The Triumph of Politics: Why thc Reagan Revolution railed New York: Harper & Row, 1986).
• A good book on thc evolution ot exchange rate arrangements in Furopc is Daniel Gros and Niels Thvgcscn European Monetary Integration: Twin the European Monetary System to Economic and Monetary Union. 2nd edn New York: Addison Wesley l.ongman 1998).
APPENDIX I: EXCHANGE RATES, INTEREST RATES AND CAPITAL SUBSTITUTABILITY
countries' bonds are to be held in equilibrium. If we define the relative risk premium attached to the home country as f>. then the interest parity relation becomes:
1 + it = (1 + 0(1 + A) iT"
The value of p as perceived by the average investor may change over time, and that is why we have given it a time subscript. If investment in the home country's bonds is perceived to be relatively riskier than in the foreign country's, the value of p will be positive (where we drop the time subscript for convenience). p will be negative if the foreign country is riskier. If their risk characteristics are identical. /> will be zero, giving the perfect substitute case, which is the standard interest parity relation. Let p> 0. Then if exchange rates are credibly fixed so that = £,, the domestic interest rate would have to exceed the foreign one to compensate for the relative risk factor.Alternatively, if interest rates are being set by their central banks in a flexible exchange rate regime, a higher value of /> will require a higher expected appreciation over time. If the future expected rate is given (as we have assumed in this chapter), the current exchange rate must immediately depreciate.This follows from rewriting the above equation as:
(1 + id
£ = . Ee
* 0 + 0(1 + Pi)
Thus, allowing for imperfect asset substitutability is straightforward. The analysis is no different from what would happen if the foreign interest rate. /*. changed. For example, a higher p (or /*) would cause the interest parity curve in Figure 20.2 to shift up. At an unchanged domestic interest rate, the exchange rate would depreciate, causing the IS curve to shift to the right, implying higher equilibrium output.
One way of thinking about p is as the relative asset demand of two countries. If /> is not zero, it means that the asset demands differ across countries for reasons other than interest rates and expected changes in exchange rates. What determines p? It will depend on the volatility of the exchange rate, the extent to which investors dislike such risk, and the extent to which the home country is a net debtor or creditor with the other country. As the home country becomes more and more in debt relative to the other country, investors will become increasingly concerned about the risk of lending to or investing in the home country. However, there may also be a subjective element driving the value of /i—international investors may decide as a group (or a herd) to change their perceptions of its true value. Thus, we have one way to think about the onset of exchange rate crises—a sudden increase in the value of p which immediately leads to a depreciation of the currency.
APPENDIX 2: FIXED EXCHANGE RATES, INTEREST RATES AND CAPITAL MOBILITY
Table 20A. I Balance sheet of the central bank
Assets Liabilities
Bonds Monetary base
Foreign-exchange reserves
Perfect capital mobility and fixed exchange rates
Consider first the effects of an open-market operation under the joint assumptions of perfect capital mobility and fixed exchange rates (the assumptions we made in the last section of this chapter).
• Assume that the domestic interest rate and the foreign interest rate are initially equal, so i = i*. Suppose that the central bank embarks on an expansionary open-market operation, buying bonds in the bond market in amount Л8. and creating money—increasing the monetary base—in exchange.This purchase of bonds leads to a decrease in the domestic interest rate. i.This is. however, only the beginning of the story:
• Now that the domestic interest rate is lower than the foreign interest rate, financial investors prefer to hold foreign bonds. To buy foreign bonds, they must first buy foreign currency. They go to the foreign-exchange market and sell domestic currency for foreign currency.
• If the central bank did nothing, the price of domestic currency would fall, and the result would be a depreciation. Under its commitment to a fixed exchange rate, the central bank cannot allow the currency to depreciate. So. it must intervene in the foreign-exchange market and sell foreign currency for domestic currency. As it sells foreign currency and buys domestic money, the monetary base decreases.
• How much foreign currency must the central bank sell! It must keep selling until the monetary base is back to where it was before its open-market operation, so the domestic interest rate is again equal to the foreign interest rate. Only then are financial investors willing to hold domestic bonds.
How long do all these steps take? Under perfect capital mobility, all this may happen within minutes or so of the original open-market operation. After these steps, the balance sheet of the central bank looks as represented in Table 20A.2. Bond holdings are up by Л8. reserves of foreign currency are down by ЛВ. and the monetary base is unchanged, having gone up by ЛВ in the open-market operation and down by A6 as a result of the sale of foreign currency in the foreign-exchange market.
To summarise: Under fixed exchange rates and perfect capital mobility, the only effect of the open-market operation is to change the composition of the central bank's balance sheet but not the monetary base.
Imperfect capital mobility and fixed exchange rates
Let's now move away from the assumption of perfect capital mobility. Suppose that it takes some time for financial investors to shift between domestic bonds and foreign bonds.
Now, an expansionary open-market operation can initially bring the domestic interest rate below the foreign interest rate. But. over time, investors shift to foreign bonds, leading to an increase in the demand for foreign currency in the foreign-exchange market. To avoid a depreciation of the domestic currency, the central bank must again stand ready to sell foreign currency and buy domestic currency. Eventually, the central bank buys enough domestic currency to offset the effects of the initial open-market operation. The monetary base is back
Table 20A.2 Balance sheet of the central bank after an open-market operation, and the induced intervention in the foreign-exchange market
Assets Liabilities
Bonds AB Monetary base ЛВ - ДВ = 0
Foreign-exchange reserves -ЛВ
to where it was before the open-market operation, and so is the interest rate. The central bank holds more domestic bonds and smaller reserves of foreign currency.
The difference between this case and the preceding one is that, by accepting a loss in foreign-exchange reserves, the central bank is now able to decrease interest rates for some time. If it takes just a few days for financial investors to adjust, the trade-off can be very unattractive—as many countries, who have suffered large losses in reserves without much effect on the interest rate, have discovered at their expense. But, if the central bank can affect the domestic interest rate for a few weeks or months, it may, in some circumstances, be willing to do so.
Now let's deviate further from perfect capital mobility. Suppose, in response to a decrease in the domestic interest rate, that financial investors are either unwilling or unable to move much of their portfolio into foreign bonds. For example, there are administrative and legal controls on financial transactions, making it either illegal or very expensive for domestic residents to invest outside the country.This is the relevant case for most middle- income countries, from Latin America, to Eastern Europe, to Asia.
After an expansionary open-market operation, the domestic interest rate decreases, making domestic bonds less attractive. Some domestic investors move into foreign bonds, selling domestic currency for foreign currency. To maintain the exchange rate, the central bank must buy domestic currency and supply foreign currency. However, the foreign-exchange intervention by the central bank may now be small compared with the initial open- market operation. And if capital controls truly prevent investors from moving into foreign bonds, there may be no need at all for such a foreign-exchange intervention.
Even leaving this extreme case aside, the net effects of the initial open-market operation and the following foreign-exchange interventions are likely to be an increase in the monetary base, a decrease in the domestic interest rate, an increase in the central bank's bond holdings, and some—but limited—loss in reserves of foreign currency. With impcrfect capital mobility, a country has some freedom to move the domestic interest rate while maintaining its exchange rate.This freedom depends primarily on three factors:
• the cegree of development of its financial markets, and how willing domestic and foreign investors are to shift between domestic assets and foreign assets
• the degree of capital controls it is able to impose on both domestic investors and foreign investors
• the amount of foreign-exchange reserves it holds: the higher the reserves, the more it can afford the loss in reserves it is likely to sustain if it decreases the interest rate at a given exchange rate.
KEYTERM
• loreign-exchange reserves, -176
CHAPTER g)
'Then 11\ agreed Until the dollar firmi up. uv let tilt elanifhell Ileal.' © The New Yorker Collection 1971. Ed Fisher, from cartoonbank.com. All rights reserved.
Exchange Rate Regimes
n July 1944. representatives of forty-four countries met in Bretton Woods, New Hampshire, in the United States, to design a new international monetary and exchange rate system. The system they adopted was based on fixed exchange rates, with all member countries other than the United States pegging the price of their currency in terms of the US dollar. In 1973, a series of exchange rate crises brought an abrupt end to the system—and an end to what is now called 'the Bretton Woods period'. Since then, the world has been characterised by many exchange rate arrangements. Some countries operate under flexible exchange rates; some operate under fixed exchange rates; some go back and forth between regimes. Which exchange rate regime is best for a country is one of the most debated issues in macroeconomics.This chapter discusses this issue.
• Section 21.1 looks at the medium run. It shows that, in contrast to the results we derived for the short run in Chapter 20, an economy ends up with the some real exchange rate and output level in the medium run, regardless of whether it operates under fixed exchange rates or flexible exchange rates. This obviously doesn't make the exchange rate regime irrelevant—the short run matters very much—but it is an important extension and qualification to our previous analysis.
• Section 21.2 looks at another aspect of fixed exchange rates, exchange rate crises. During a typical exchange rate crisis, a country operating under a fixed exchange rate is forced, often under dramatic conditions, to abandon its parity and to devalue. Such crises were behind the breakdown of the Bretton Woods system. They rocked the European Monetary System in the early 1990s, and were a major element of the Asian financial crisis of the late 1990s. It is important to understand why crises happen and what they imply.
• Section 21.3 turns to the behaviour of exchange rates under a flexible exchange rate regime. It shows that the behaviour of exchange rates and the relation of the exchange rate to monetary policy are in feet more complex than we assumed in Chapter 20. Large fluctuations in the exchange rate, and the difficulty of using monetary policy to affect the exchange rate, make a flexible exchange rate regime less attractive than it appeared to be in Chapter 20. The section ends with an integrated analysis of the dynamics between the short run and the medium run in a flexible exchange rate regime.
• Section 21.4 puts these results together, reviewing the case for flexible or fixed rates. It discusses two recent and important developments. The first was the move to a common currency in Europe. The case for the euro is used to introduce the arguments for and against a common currency between Australia and New Zealand. The second development is the move towards strong forms of fixed exchange rate regimes, from currency boards to dollarisation.
21.1 FIXED EXCHANGE RATES AND THE ADJUSTMENT OF THE REAL EXCHANGE RATE IN THE MEDIUM RUN
The results wc derived in Chapter 20, where we loeused on the short run, drew a sharp contrast between the behaviour ot the economy under flexible exchange rates and under fixed exchange rates.
• Under flexible exchange rates, a country that needed to achieve a real depreciation—tor example to reduce its trade deficit or to get out ot a recession—could do so by using monetary policy to decrease the interest rate and decrease the exchange rate achieve a depreciation .
• Under fixed exchange rates, a country lost both ol these instruments. 13y definition its nominal exchange rate was lixed and thus couldn't be adjusted. And the fixed exchange rate and thc interest parity condition implied lhal ihe country couldn't adjust its interest rate.- the domestic interest rate had to remain equal to the foreign interest rate.
This appeared to make a flexible exchange rate regime much more attractive than a fixed exchange rate regime. Why give up two macroeconomic instalments: As we now shilt focus Irom the short atn to the medium run, you will see that this earlier conclusion needs lo be qualilied. While the conclusions about the short am were valid, you will see that in the medium am the difference between thc two regimes fades away. More specifically, in the medium run the economy reaches the same real exchange rale and the same level of output whether it operates under fixed exchange rates or flexible exchange rates.
Thc intuition lor this result is straightlorward. Recall the definition ot the real exchange rate:
LP
I he real exchange rate, e, is equal to the nominal exchange rate. F. (the price of domestic currency in terms ot foreign currency, times the domestic price level. P. divided by the foreign price level, P'. There arc therefore, two ways in which the real exchange rate can adjust:
• I hrough a change in the nominal exchange rate, E. This can be done only under llexiblc exchange rates. And il we assume that the domestic price level P. and the foreign price level, P don t change in the short am, il is the only way lo adjust ihe real exchange rale in thc short run.
• Through a change in thc domestic pricc level, P. relative to thc foreign pricc level. P In the medium run this option is open even to a country operating under a fixed ( nominal) exchange rate. And ihis is indeed what happens under fixed exchange rates. The adjustment takes place through the price level rather than through the nominal exchange rate.
Warning:The next paragraphs rely on what
you learned in earlier ^ chapters. Make sure you remember the definitions of the real interest rate (Chapter 14). the real exchange rate (Chapter 18) and the interest rate panty condition (Chapter 18).
Let us go through this argument sicp by step. To begin, lei us derive the aggregate demand and aggregate supply relations tor an open cconomy under fixed exchange rates. Later, in Section 2 1.3, we will build the flexible exchange rate analogue.
Aggregate demand under fixed exchange rates
Start from thc condition lor goods-market equilibrium derived in Chapter 20. equation (20.1):
Y=C(Y-T) + l(Y,r) + G -1- NX{Y,Y',e) (21.1)
I his condition states that, lor thc goods market to be in equilibrium, output must be equal to thc demand lor domestic goods—that is, the sum ol consumption, investment, government spending and net exports.
r = i - Tf EP
e = T
Ncxi rccall thc following relations: • The real interest rate, r. equals ihe nominal interest rate. /', minus expected inflation, 77':
The real exchange rate, e is defined as
Under fixed exchange rates, the nominal exchange rale, E, is. by delinition, fixed. Denote by E lhe value at which lhe nominal exchange rate is lixed, so
E - Ё
Under fixed exchange rates and perfect capital mobility and substitutability, the domestic interest rate, i, must be equal to the lorcign interest rate, i":
Using these lour relations, rewrite equation (21.1 as
EP P'
21.2
Y = ОУ - T) - KY,i* - v*) + G + NXIY.Y
This is a rich—and complicated—equilibrium condition. It lells us that in an open economy with fixed exchange rates, equilibrium output or. more precisely the level ol output implied by equilibrium in lhe goods, linanciai and torcign-cxchange markets) depends on:
• Government spending, G, and taxes, T. An increase in government spending increases output. So docs a decrease in taxes.
• The foreign nominal interest rale, Г minus expected inflation, тг'. An increase in thc foreign nominal interest rate requires a parallel increase in the domestic nominal interest rate. Given expected inflation, this increase in the domestic nominal interest rate leads lo an increase in the domestic real interest rate and so decreases demand and output.
• foreign output, Y*. An increase in foreign output increases exports and so increases net exports. 7 he increase in net exports increases domestic output.
• The real exchange rate e equal to the fixed nominal exchange rate. E. times the domestic price level. P, divided by the foreign price level, P . A decrease in the real exchange rate—equivalently, a real depreciation—leads to an increase in net exports and so to an increase in output.
(21.3)
Wc will locus here on the effects ol only three ol these variables: the real exchange rate, government spending and taxes. Let us write thc relation between these three variables and output as
VI y.G.T
(
A decrease in thc real exchange rate—a real depreciation—increases output. So does an increase in government spending or a decrease in taxes. All the other variables that allcct output in equation 21.2 are taken as given, and to simplify notation we simply omit them Irom equation (21.3).
Equation i21.3i gives us our aggregate demand relation the relation between output and the price level implied by equilibrium in the goods market and in financial markets. As in the closed economy this aggregate demand relation implies a negative relation between the price level and output. But, while the sign of the effect ol the price level on output in the aggregate demand relation is the same as in thc closed economy, the channel is very dilfercnt:
• In the closed economy, the price level allects output through its effect on monetary policy, leading to a change in interest rate.
In the closed economy, we didn't need the LM relation to derive the aggregate demand relation. The reason is that the nominal interest rate was set by the central bank on the basis of its monetary policy rule. Under fixed exchange rates, again we don't need the LAI relation.The interest rate here is determined by (must be equal to) the foreign interest rate. (In both the closed economy and the open economy, the LM
* relation still holds, but, as we saw in Chapters 5 and 20. it simply determines the money stock.)
4 In a closed economy.
/»T=>(P Pr)t=>it=>yi
In an open economy, in addition
PT^£P/P*T->NXJ.=>rl
• In the open economy under fixed exchange rates, the interest rate is lixed—pinned down by the foreign interest rate. The way the price level affects output here is instead, through its effect on the real exchange rate. Given the fixed nominal exchange rate. E. and the foreign price level, P ', an increase in the domestic pricc level, l\ leads to an increase in the real exchange rate Г.Р/Р -a real appreciation. This real appreciation leads to a decrease in net exports, and a decrease in demand and in turn, to a decrease in output. Put simply, an increase in the price level makes domestic goods more expensive, thus decreasing thc demand for domestic goods and in turn decreasing output
Equilibrium in the short run and the medium run under fixed exchange rates
Thc aggregate demand curve under fixed exchange rates implied by equation (2l.3i is drawn as thc AD curve in Figure 21.1. It is downward sloping: an increase in thc pricc level decreases output. As always, the relation is drawn lor given values ol all other variables in this case, lor given values of £. P~, G and T.
For thc aggregate supply curve, we rely on the relation derived in thc core. Going back to the aggregate supply relation derived in Chapter 7, equation '7.2):
P - P'\l + /i)F^I-y,zj (21.4)
Thc price level, P, depends on the expected price level, P and thc level of output, V. Recall the two mechanisms at work:
P ' :W _iP* ► • The expected price level matters because it allccts nominal wages, which in turn affcct the price level
• 1 ligher output matters because it leads to higher employment, which leads lo lower unemployment, V —>:ii--> wT-->/'* ► which leads to higher wages, which lead to a higher price level.
The aggregate supply curve is drawn as thc /IS curve in Figure 21.1 for a given value of the expected pricc level. It is upward sloping: higher output leads to a higher pricc level.
The short-run equilibrium is given by the intersection ot the aggregate demand curvc and the aggregate supply curve, point A in Figure 21.1. As was the case in the closed economy, there is 110 reason why the short-run equilibrium level ot output, Y, should be equal to thc natural level ol output, Y„. As the figure is drawn, У is lower than Y,„ so output is below the natural level ol output.
What happens over lime? The basic answer is familiar from our earlier study ot adjustment in a closed economy, and is shown in Figure 21.2. So long as output remains below the natural level ol
Figure 21.1 Aggregate demand and aggregate supply in an open economy under fixed exchange rates
Y
Output, Y
An increase in the price level leads to a reai appreciation and a decrease in output: t)>e aggregate demand curve is downward sloping. An increase in output leads to an increase in the price level: the aggregate supply curve is upward sloping.
У
Output, У
Figure 21.2 Adjustment under fixed exchange rates
Suppose that the government decides, while keeping the fixed exchange rate regime, to allow for a one-time devaluation. For a given price level, a devaluation (a decrease in the nominal exchange rale) leads to a real depreciation a decrease in the real exchange rate), and thus to an increase in output. In other words, a devaluation shifts ihe aggregate demand curve to the right: output is higher at a given price level.
This has a straightforward implication-, a devaluation of the right size can take the cconomy directly from У to Y„. This is shown in Figure 21.3. Suppose the economy is initially at A. thc same point .4 as in Figure 21.2. Thc right size devaluation shifts the aggregate demand curve from AD to AD', taking the equilibrium from A to C. At C, output is equal to the natural level of output Y„. The real exchange rate is thc same as at B. (We know this becausc output is thc same at points В and C. From equation [21.3], and without changes in С or T. this implies that the real exchange rate must also be the same.)
That the right size devaluation can return output to thc natural level of output right away—rather than over time, as was the case in the absence of the devaluation—sounds too good to be true and, in practice, it is. Achieving the right size devaluation—the devaluation that takes output to Y„ right away—is easier to achieve in a graph than in reality:
• In contrast to our simple aggregate demand relation 21.3. thc effects of the devaluation on output don . happen right away. As you saw in Chapter 19. the initial effects of a devaluation on output may be contractionary as people pay more for imports, and thc quantities ol imports and exports haven't See Section 19.5 on the ► yet adjusted.
-curve . Also, in contrast to our simple aggregate supply relation (21.4), there is likely to be a direct effect ol the devaluation on the price level. As the price ol imported goods increases the price of a consumption basket increases. This increase is likely to lead workers to ask lor higher nominal wages, forcing firms to increase their prices as well.
AD
Hut these complications don't affect the basic conclusion: allowing thc nominal exchange rate to adjust can help output return to its natural level, il not right away, at least faster than without a
Figure 21.3 Adjustment with a devaluation
У Yn
Output, Y
The right size devaluation can shift aggregate demand to the right, leading tile economy to point C. At point C, output is back to the natural level of output.
devaluation. And so. whenever a country under lixed exchange rates laces either a large trade delicit or a large recession, there is в lot ol political pressure either lo give up the lixed exchange rate regime altogether or. at least, to have a one-time devaluation. Perhaps the most lorcetul presentation of this view was made nearly eighty years ago by Keynes, who argued against Winston Churchill's decision to return thc British pound in ; 025 lo its pre-World War I parity. I lis arguments are presented in thc focus box The return of Britain to the gold standard-. Keynes versus Churchill'. Most economic historians believe thai history proved Keynes right, and that overvaluation of the pound was one of the main reasons lor Britain's poor economic performance after World War I.
racu:
'BOX
Those who oppose a shift to flexible exchange rates or a devaluation argue that there are good reasons to choose fixed exchange rates, and that loo much willingness to devalue deleats the purpose
THE RETURN OF BRITAIN TO THE GOLD STANDARD: KEYNES
Vi 3; VI > a I'J;
In 1925 Britain decided to return to the gold standard. The gold standard was a system in which each country fixed the price of its currency in terms of gold and stood ready to exchange gold for currency at the stated parity. This system implied fixed nominal exchange rates between countries.
The gold standard had been in place from 1870 until World War I. Because of the need to finance the war, and to do so in part by money creation, Britain suspended the gold standard in 1914. In 1925 Winston Churchill, then Britain's chancellor of the exchequer (the British equivalent of the treasurer in Australia), decided to return to the gold standard, and to return at the prewar parity—that is, at the prewar value of the pound in terms of gold. But. because prices had increased faster in Britain than in many of its trading partners, returning to the prewar parity implied a large real appreciation. At the same nominal exchange rate as before the war. British goods were now more expensive relative to foreign goods. (Go back to the definition of the real exchange rate, e = EPiP : the price level in Britain, P. had increased more than the foreign price level, P . At a given nominal exchange rate, £, this implied that e was higher, that Britain suffered from a real appreciation.)
Keynes severely criticised the decision to return to the prewar parity. In The Economic Consequences of Mr Churchill, a book he published in 1925, Keynes argued as follows: if Britain was going to return to the gold standard, it should have done so at a higher price of gold in terms of currency, at a nominal exchange rate lower than the prewar nominal exchange rate. In an article that appeared in The Nation and Athenaeum newspaper on 2 May 1925, he articulated his views as follows:
There remains, however, the objection to which I have never ceased to attach importance, against the return to gold in actual present conditions, in view of the possible consequences on the state of trade and employment. I bel eve that our price level is too high, if it is converted to gold at the par of exchange, in relation to gold prices elsewhere; and if we consider the prices of those articles only which are not the subject of international trade, and of services, i.e., wages, we shall find that these are materially too high— not less than 5 per cent, and probably 10 per cent.Thus, unless the situation is saved by a rise of prices elsewhere, the Chancellor is committing us to a policy of forcing down money wages by perhaps 2 shillings in the pound.
I do not believe that this can be achieved without the gravest danger to industrial profits and industrial peace. I would much rather leave the gold value of our currency where it was some months ago than embark on a struggle with every trade union in the country to reduce money wages. It seems wiser and simpler and saner to leave the currency to find its own level for some time longer rather than force a situation where employers are faced with the alternative of closing down or of lowering wages, cost what the struggle may.
For this reason. I remain of the opinion that the Chancellor of the Exchequer has done an ill-judged thing—III judged because we are running the risk for no adequate reward if all goes well. Keynes's prediction turned out to be right. While other countries were growing. Britain was in recession for the rest of the decade. Most economic historians attribute a good part of the blame to the initial overvaluation.
Because it is more ► convenient, we use the approximation, equation (18.4), rattier than the original interest parity condition, equation (18.2). We will usually use the approximate form when giving numerical examples, simply because it gives round numbers. But be aware of the dangers of this approximation as interest rates increase.
•21.5'
1, -
As an exercise, calcu ate these numbers using the exact form of the interest parity relation given in equation (20.4a). How far from1 60 per cent is the needed interest rate differential?
of adopting a fixed exchange rale regime in the lirst place. I hev argue lhat too much willingness on the part of governments to consider devaluations actually leads to an increased likelihood ol exchange rate crises. To understand their arguments, we now turn to these crises, what triggers them and what their implications might be.
21.2 EXCHANGE RATE CRISES UNDER FIXED EXCHANGE RATES
What, then, arc the choices confronting the government and the central bank?
• l irst, they can try to convince markets that they have no intention of devaluing. This is always the lirst line ol delence. Communiques are issued, and prime ministers or presidents appear on TV to reiterate their absolute commitment to the existing parity. But words are cheap, and they rarely convince linancial investors.
• Second, the central bank can increase the interest rate, but by less than would be needed to satisly equation 21.5)—in our example, by less than 60 per cent. Although domestic interest rates are high, thev aren't high enough to fully compensate tor the perceived risk of devaluation. This action typically leads to a large capital outllow, as financial investors still prefer to get out ol domestic bonds and into foreign bonds. This implies selling domestic bonds, getting the proceeds in domestic currency, going to thc foreign-exchange market to sell domestic currency (or foreign currency, and then buying foreign bonds. It the central bank didn't intervene in the foreign-exchange market the large sales ol domestic currency for foreign currency would lead to a depreciation. If it wants to maintain the exchange rate, the central bank must therefore stand ready to buy domestic currency and sell foreign currency at thc current exchange rate In doing so, it often loses most of its reserves of foreign currency. (The mechanics of central hank intervention were described in the second appendix to Chaptcr 20.)
• Eventually—after a few hours or a few months—the choice for the central bank becomes either to increase the interest rate enough to satisfy equation 21.5) or to validate the market's expectations and devalue. Setting a very high short-term domestic interest rate can have a devastating effect on demand and on output. This course ot action makes sense only if (1) the perceived probability ol a devaluation is small, so the interest rate doesn't have to be too high, and '2) the government believes that markets will soon become convinced that no devaluation is coming, allowing domestic interest rates to decrease. Otherwise, the only option is to devalue.
To summarise.- Expectations that a devaluation may be coming can trigger a:i exchange rate crisis. Faced with such expectations, the government has two options: I > give in and devalue, or' 2) tight and maintain the parity, at the cost of very high interest rates and a potential recession. Fighting may not work anyway- the recession may force the government to change policy later on, or force the govern¬ment out of office.
See the focus box in Chapter 5. German unification and the German monetary-fiscal tug of war', and the focus box in Chapter 20.
* 'German unification, interest rates and the EMS'
An interesting twist here is that a devaluation may happen even il the belief that a devaluation was coming was initially groundless. Even il the government initially had no intention ol devaluing, it may be forced to devalue if financial markets believe that it will devalue. The cost ol maintaining the parity would be a long period ol high interest rates and a recession,- thc government prefers to devalue instead.
The 1992 EMS crisis
An example ot the problems discussed in this section is the exchange rate crisis that shook the European Monetary System EMS) in the early 1990s.
At the start ot thc 1990s, thc EMS appeared to be working well. Started in 1979, it was an exchange rate system based on lixed parities with bands. Each member country (among them France, Germany, Italy and starting in 1990, the United Kingdom had to maintain its exchange rate against all other member countries within narrow bands. Thc lirst few years had been rocky, with many realignments— adjustments ol parities—among member countries, but from 1987 to 1992 there were only two realign¬ments. There was increasing talk about narrowing thc bands further and even moving to the next stage—to a common currency.
In most countries, the
^ government is formally in charge of choosing the parity, and the central bank is formally in charge of maintaining it In practice, choosing and maintaining the parity are joint responsibilities of the government and the central bank.
4 In the summer of 1998. Boris Yeltsin announced that the Russian government had no intention of devaluing the ruble.Two weeks later, the ruble collapsed.
In 1992, however, financial markets became increasingly convinced that more realignments were soon lo come. The reason was one we have seen already—namely, thc macroeconomic implications of German reunification. Because of the pressure on demand coming from reunification, the Bundesbank (the German central bank' was maintaining high interest rates to avoid too large an increase in output and an increase in inflation in Germany. While Germany's EMS partners needed lower interest rates to reduce growing unemployment, they had to match thc German interest rates to maintain their EMS parities. To financial markets, the position of Germany's EMS partners looked increasingly untenable.
Lower interest rates outside Germany, and thus devaluations of many currencies in relation to the deutschmark, appeared increasingly likely.
Throughout 1992, the perceived probability ol a devaluation forced several of Germany's trading partners to maintain higher nominal interest rates than Germany. Hut the lirst major crisis didn't come until September 1992. The day-by-day story is tolcl in the focus box Anatomy of a crisis: The September 1992 EMS crisis . The belie! that several countries were soon going to devalue led in early September to speculative attacks on several currencies with linanciai investors selling in anticipation ol an oncoming devaluation. All the lines ol defence described earlier were used by the monetary authorities and the governments ol thc countries under attack. First, solemn communique's were issued, but with no discernible effect. Then, interest rates were increased, up to 500 per cent lor the overnight interest rate (the rate lor lending and borrowing overnight in Sweden (expressed at an annual rate . But they weren't increased enough to prevent capital outflows and large losses ol foreign-exchange reserves by the central banks under pressure. Next came different courses ol action in dillerent countries: Spain devalued its exchange rate. Italy and the United Kingdom suspended their participation in the FMS, and France decided to tough it out through higher interest rates until the storm was over.
By thc end ot September financial markets believed that no further devaluations were imminent. Some countries were no longer in the EMS, others had devalued hut remained in the EMS and those that had maintained their parity had shown their determination to stay in the EMS, even if this meant very high interest rates. But thc underlying problem—thc high German interest rates—was still present, and it was only a matter of time until the next crisis. In November 1992, further speculation forced a devaluation ot thc Spanish peseta, thc Portuguese cscudo and thc Swedish krona. The peseta and thc escudo were further devalued in May 1993. In July 1993, after yet another large speculative attack, EMS countries decided to adopt large fluctuation bands plus or minus 15 per cent) around central parities, in ellect moving to a system that allowed for very large exchange rate fluctuations. This system with wider bands was kept until the adoption of a common currency in January 1999.
us
FQC 1
To summarise: The 1992 LMS crisis came Irom the perception by financial markets that the high interest rales lorced by Germany on its partners under lhe niies ol the EMS were becoming very costly. The belief lhat some countries might want to devalue or get out of the EMS led investors to ask lor even higher interest rates making ii even more costly lor those countries lo maintain their parity. In the end, some countries couldn't bear the cost, some devalued, some dropped out. Others remained in the system, but at a substantial cost in terms ot output.
E SEPTEMBER 1992 EMS CRISIS
previous devaluation) the lira. Both England and Italy announce that they are temporarily suspending their participation in the ERM. Over the following weeks, both currencies depreciate by roughly 15 per cent against the DM.
• September 16-17: With the pound and the lira out of the ERM, the attack turns against the other currencies.To maintain its parity, Sweden increases its overnight rate to 500 per cent! Ireland increases its overnight rate to 300 per cent. Spain decides to stay in the ERM. but to devalue by 5 per cent.
• September 20: French voters narrowly approve the Maastricht Treaty (the treaty that sets the timetable for the transition to a common currency) in a referendum. A negative vote would surely have amplified the crisis.The narrow, but positive, vote is seen as the sign that the worst may be over, and that the treaty will eventually be accepted by all EU members.
• September 23-28: Speculation against the franc forces the Banque de France to increase its short-term interest rate by 2.S per cent. To defend their parity without having to resort to very high short-term interest rates, both Ireland and Spain reintroduce capital controls.
• End of September:The crisis ends.Two countries, the United Kingdom and Italy, have left the ERM and let their currency depreciate. Spain remains within the ERM. but only after a devaluation.The other countries have maintained their parity, but, for some of them, at the cost of large reserve losses.
i See the relation becween the two in Figure 20.1.
SOURCE:World Economic Outlook. October 1993.
21.3 EXCHANGE RATE MOVEMENTS UNDER FLEXIBLE EXCHANGE RATES
In the model we developed in Chapter 20, there was a simple relation between the interest rale and lhe exchange rate: ihe lower thc interest raic. the lower the exchange rate. This implied that a country that wanted to maintain a stable exchange rate just had to maintain its interest rate close to the foreign interest rate. A country that wanted to achieve a given depreciation iust had to decrease its interest rate by the right amount.
In reality, the relation between the interest rate and the exchange raic isn't so simple. Exchange rates ollen move even without movements in interest rates These exchange rate moves have real effects on the economy—output and prices will change, thus allecting monetary policy. The size of the effect of a given decrease in the interest rate on thc exchange rate is hard to predict, making it much harder for monetary policy to achieve its desired outcome
4 To derive what fol ows. we could just as easily use the approximate version of interest parity, equation (21.5).
-
To see why things arc more complicated, we must return once again to the exact interest parity condition derived in Chapter 18 equation 18.2 :
i + i, - £,( i + >",)
ЕЧ.
Rewrite it as
(2I.6-
Think of the time period i from I to I + 11 as one year, l he exchange rate ihis year depends on the one-year domestic interest rate, the one-year foreign interest raic and thc exchange rate expected for next year. Wc assumed in Chapter 20 that ihe expected exchange rate next year (E,+l > was constant. But this was a simplification The exchange rate expected one year hence isn't constant. Using equation 12 1.6), but now lor next year, it is clear lhat the exchange rale next year will depend on next year's onc- vcar domestic interest rate the one-year lorcign interest rate and the exchange rate expected lor the year alter, and so on. So. any change in expectations ol current and luture domestic and foreign interest rates, as well as changes in the expected exchange rate in the far future, will affect the exchange rate- today.
Lot's explore this more closely. Write equation (2 1.61 tor year t + I rather than year t■.
F - ' + F' 4 I - ,, % Cl-2
С +
_ ' * 'Т.. £,
The expected exchange rate in year / • I depends on the domestic interest rate expected for year / + I, the foreign interest rate expected lor year ( + I, and the expected future exchange rate in year t ■ 1. Replacing £'.'., in equation ( 2 1.6) gives
(1 TQ(1 + &,)
The current exchange rate depends on both this years and next years expected domestic and foreign interest rates, and on the expected exchange rate two years from now. Continuing to solve forward in time in the same way bv replacing C'..2, Land so on until, say, year t + n), we get
11 + /,)(i + /','.i)... (I w7.,_,)
fc| - (l + oo+ft.b-.o+Sv.)fc"" {2U7)
Suppose we take n to be large—say, ten years. (Equation I21.7J holds for any value of n. • This relation tells us that the current exchange rate depends on two sets of factors:
• current and expected domestic and foreign interest rates for each year over thc next ten years
• the expected exchange rate ten years from now.
For some purposes, it is useful to go further and derive a relation between current and expected future domestic and loreign real interest rates, the current real exchange rate and the expected future real exchange rate. This is done in the appendix to this chapter. The derivation isn't much fun, but it is a useful way of brushing up on the relation between real interest rates and nominal interest rates, and real exchange rates and nominal exchange rates. Equation (21.7; is sufficient, however, lo make the three poinis we want to emphasise here.
Exchange rates and the current account
Any lactor that moves the expected future exchange rate. £,.,„ moves the currcnt exchange rate, £,. Indeed, il ihe domestic interest rate and the foreign interest rate arc expected to be thc same in both countries trom t to t + it. the traction on the right in equation <21.71 is equal to one, so the relation reduces lo E. = £','+„. Thc effect ol any change in the expected future exchange rate on the current exchange rate is one lor one.
It we think of n as large (say. icn years or more, wc can think of £,.„ as thc exchange rate required to achieve current account balance in the medium or long run. Countries cannot borrow—mn a current account delicit—forever, and will not want lo lend—run a current account surplus—forever, either. Thus, any news that alfccts forecasts ot the current account balancc in thc luture is likely to have an effect on the expected future exchange rate and, in turn, on ihe exchange rate today. For example, thc announcement ol a larger-than cxpectcd trade deficit may lead investors to conclude thai a depreciation will be needed at some point to re-establish trade balancc. Thus, £•;.„ will decrease, leading in turn lo a decrease in £, today.
One critical lacior that has a persistent effect on the Australian current account and thus the Australian dollar is world commodity prices. As you saw in Chapter 18. 65 per cent of Australia's
exports today are mining products (mostly coal) and I 3 per cent are rural products. Therefore the world prices of these commodities are going to have a serious effect on Australian export revenue, and in turn the current account. This means that the exchange rate should strengthen when commodity prices rise, figure 21.4 presents the data lor world commodity prices and Australia's nominal effective exchange rate. There is clear relationship between these two series. Thc Australian dollar strengthened significantly Irom 2002 as world commodity prices surged. However, as you saw in Figure I S.K, the current account has remained in serious deficit '0 per cent ol GDP in 2008), and so when the global financial crisis became acute in late 2008, thc Australian dollar fell dramatically, in spite ol still strong world commodity prices. One possible reason lor the big fall in the Australian dollar is that world commodity prices arc expected to continue weakening in 2009. And wc know Irom equation 21.7 that expectations such as these can cause major changes in the currcnt exchange rate.
Exchange rates and current and future interest rates
Any factor that moves current or expected future domestic or foreign interest rates between year I and t + n moves the current exchange rate, for example, given foreign interest rates, an increase in current or expected luture domestic interest rates leads to an increase in F, so to an appreciation.
4 See Chapters 18 and 20.
For more on the relation between long- term interest rates and * current and expected future short-term interest rates, go back to Chapter 15.
This implies that any variable that leads investors to change their expectations ol future interest rates will lead to a change in the exchange rale today. For example, the dance ol the US dollar in the 1980s and of thc Australian dollar and thc yen in thc 1990s discussed earlier can he largely explained by the movement in current and expected luture domestic interest rates relative lo interest rates in the rest ol the world during lhat period. So, lor example, at the end of the 1980s, light monetary policy raised Australian short-term and long-term interest rates, with the increase in long-term rates reflecting anticipations ol high short-term interest rates in the luture. This increase in both current and expected future interest rates was, in turn, the main cause of the Australian dollar appreciation. Both liscal policy and monetary policy were reversed in 1991, leading to lower Australian interest rates and an Australian dollar depreciation.
Г 85 Figure 21.4
World
-80 commodity prices
and the Australian
-75 dollar. 1984-2008
-70 m
f
о
-65 Ф
-60 a
X n
3-
-55 5
к го
-50 ■г
■
rt
a
-45
-40
-35
280 -,
240
Я 200
160
120
80 -
Nominal effective exchange rate right scale
World commodity price index- left scale
in in и inn II in пит II HI II i II II i mi и i II II n i II HI и и i II in и
1984 1986 1983 1990 1992 1994 1996 1998 2000 2002 2004 2006
x 01 -o с
a £
'■5 о E
£ о
I
40
There is a significantly positive relationship between movements in world commodity price Iin SDR terms) and the Australian dollor /effective nominal exchange rate).
SOURCE. RBA BuWern.Tables GS. Fl I
Exchange rate volatility
The third implication follows Irom the lirst two In reality, and in contrast to our analysis in Chapter 20, thc relation between the interest rate, /,. and the exchange rate, £,, is anything but mechanical. When the central bank cuts thc interest rate, financial markets have to assess whether this action signals a major shilt in monetary policy and the cut in thc interest rate is just the first ol many such cuts, or whether this cut is just a temporary movement in interest rates. Announcements by the central bank may not be very useful. The central bank itself may not even know what it will do in the future. Typically, it will be reacting to early signals, which may be reversed later. Financial markets also have to assess how foreign central banks will react whether they will stay put or follow suit and cut their interest rates. All this makes it much harder to predict what the effect of thc change in the interest rate- will be on thc exchange rate.
Let's be more concrete. Go back to equation (21.7). Assume that £','.„ - I. Assume that currcnt and cxpcctcd future domestic interest rates and current and expected future foreign interest rates are all equal lo 5 per cent. Thc current exchange rate is then given by
£, =
I
(1 0.05)" (1 • 0.05)"
Now consider a monetary expansion, which decreases the current domestic interest raic. from 5 per cent lo 3 per cent. Will ihis lead lo a decrease in C,—to a depreciation—and it so, by how much? The answer: It all depends.
1.03 1.05
E. =
0,98
I =
Suppose the interest rate is expected to be lower just for one year, so the 11 - I expected future interest rales are unchanged. The current exchange rale then decreases to
s i" i
1.03(1 + 0.05) (1 + 0.05)"
The expansionary moneiarv policy leads to a decrease in ihe exchange rate—a depreciation—of only 2 per cent.
(1.03)5(l + 0.05)
Е/
(1.05)
Suppose instead that when the current interest rate declines Irom 5 per cent lo 3 per cent, investors expect the decline to last for five years (so, i.., - .
/.. | /, - 3%). The exchange rate then decreases to "5 (I.03)5
= 0.91
i 1 + 0.05)"
The expansionary monetary policy now leads to a decrease in the exchange rate—a depreciation— of 9 per cent, a much larger effect.
You can surely think of vet other outcomes. Suppose investors had anticipated that the ccniral bank was going ю decrease interest rates, and the actual decrease turns out to be smaller than they anticipated. They will revise their expectations ol future nominal interest rates upward, leading to an appreciation rather than a depreciation of thc currency!
When at the end ot the Bretton Woods period, countries moved Irom fixed exchange rales lo flexible exchange rates, most economists had expected that exchange rates would be stable. Thc large fluctuations in exchange rates thai followed (and have continued to this day) came as a surprise. For some time, these fluctuations were thought to be thc result of irrational speculation in foreign-exchange markets. It wasn't until thc mid-1970s thai economists realised that these large movements could be explained, as we have- here, by the rational reaction of financial markets to news about future interest rates and the future exchange rate. This has an important implication: a coumrv thai dccides to operate under flexible exchange rates must accept thc lact thai it will be exposed to substantial exchange rate fluctuations over time.
Overshooting exchange rates
If this reminds you of ► our discussion on the effect of monetary policy on stock prices in Chapter 15, you are right.This is more than а coincidence. Like stock prices, the exchange rate depends very much on expectations of variables far into the future. How expectations change in response to a change in a current variable (here, the interest rate) very much determines the
In thc previous subsection, you saw how thc current exchange rates depended on current interest rates and expectations ot future interest rates, and thus monetary policy. It isn't a trivial matter ю determine
these expectations. Hardly a day goes hy in the media without the luture intentions ol thc central hank being discussed. Central banks may publicise their objectives (lor example, the RBA has a declared inflation target >, but they never commit themselves to a hard and last rule. When we introduced interest rate rules lor thc central bank in Chapter 7 'equation [7.31; and Chapter 9 (equation [9.91), this was our simple way ol modelling the central banks behaviour. But their behaviour is never so simple. I here arc- so many and varied shocks that allect economies, and central banks reserve the right to respond as they see fit. Their actions often surprise the market:
Nevertheless, to improve our understanding ol thc behaviour of the exchange rate, let us lirst think about a blatantly unrealistic case. Assume that there is no inllation, here or abroad, current or expected, so that we don't need to distinguish between nominal and real interest rates or between nominal and real exchange rates. Suppose further that, initially, domestic and foreign interest rates are expected to he constant and equal to each other, and the economy is in a medium-run equilibrium.
Now suppose that the domestic central bank unexpectedly announces that, to decrease economic activity, it has decided to increase thc interest rale, so lhat it will be 2 per cent higher lor lhe next live- years, after which it will return to normal. Financial markets fully believe this announcement.
What is the effect on the exchange rate today? To answer this question, we work backwards in time.
• Start five years in the future I he exchange rate five years Irom now depends on what is expected to happen to interest rates thereafter, as well as on the medium-run exchange rate. Because the announcement doesn't change expectations ol interest rates beyond the five years, and presumably doesn't change the mcd um-run exchange rate either, there is no change in the expected exchange rate live years Irom now.
• What happens between today and live years hence? Think about it this way. Because it gives round numbers, we will use the approximate interest parity condition, equation (20.4c . Il tells us lhat for each of the next five years there must be an expected depreciation of the domestic currency of 2 per cent per year, so thai the expected rates ol return on holding domestic and loreign bonds are equal. Thus, there must be an expected cumulative depreciation ol 5 X2% = 10% over the next five years. As the expected exchange rate live vears hence is unchanged from what it was before the announce¬ment. there must be an immediate appreciation today of 10 per cent, so as lo generate lhe expected depreciation of 10 per cent over thc next five years. In other words, if the domestic currency appreciates by 10 per cent today and is then expected to depreciate by 2 per cent per year for the next live years, linanciai investors will be willing to hold foreign bonds, even though the domestic: interest rale is 2 percent higher than the foreign interest rate.
The expected paths ol the interest rate and the exchange rale are shown in 1 igure 21.5. Before the announcement, the domestic interest rale is /,, and the exchange rate is E,,.These are their medium-run values. For the next five years, the interest rate is raised to /„ > 2%. The exchange rate immediately appreciates by 10 percent to E(] (I + 10%!. Thereafter, il depreciates by 2 percent per year—by year I, it reaches E,, I + 8%), and so on. Alter live years, the- interest rate and the exchange rate arc back ai their pre-announcement values—their medium-run values. Note how much the exchange rate suddenly moves, overshooting its medium-run value—increasing first, only to decrease back to its initial value- live years later For that reason, this exchange rale adjustment is often referred to as overshooting.
Let us now sec this phenomenon at work in our open economy macroeconomic model under flexible- exchange rates.
Equilibrium in the short run and medium run under flexible exchange rates
4 Have a go at answering question 7 at the end of this chapter for more intuition of the overshooting result.
We can now build on what we learned in previous chapters to tell a more complete story ol the behaviour of exchange rates in the short run and medium run. The easiest way to do this is to take thc- closed economy mode', ol Chapter 7 and add our open economy extensions. In Chapter 7 we showed what would happen in a closed economy under thc convenient assumption lhat the central bank used the interest rate as its instrument to achieve a price level target. This meant that inllation was zero in lhe medium run, which is a great simplifies Though inllation may occur in the short run, we will ignore that and thus make no distinction between nominal and real interest rates.
Years
Figure 2I.S The effects of a monetary policy change on the exchange rate
Output, У
Eo
Exchange rate, E
Interest parity0
Figure 21.6 The open economy macro model under flexible exchange rates
The real exchange rate will only change in the medium run if there are real shocks. For ^ example, a fiscal contraction will require an increase in net exports to compensate, and so the real exchange rate would depreciate.
(c)
Output, Y
Output, У
Eo E,
Exchange rate, E
(c)
Figure 21.7 The medium-run effects of a monetary policy contraction under flexible exchange rates
V„
Output, У
Monetary policy contraction—from the short run to the medium run
Now we are ready to explain the short-ntn equilibrium effects in the lirst year alter the new policy has heen introduced, and the transition to the new medium-run equilibrium. We will describe the effects only briefly here.
The short-run equilibrium after the monetary policy contraction is at points R in Figure 21.8. As soon as the central bank decides on a lower target. Pj. it begins by increasing the interest rate to i|t. This requires open-market operations to shift the /..VI curve up to /Л1ц. The lower P' (and higher P) also leads to a leltward shilt ol the AD curve lo (and IS to IS/tl. lust as you saw in the closed economy model ol Chapter 7. raising thc interest rate generates a recession in the short run—output lalls. But there is an additional reason why output lalls in lhe open economy in the first year. The leltward shift
LMo
Output, Y
(b)
IP о
to E,
Exchange rate, E
о О
Figure 21.8 The short-run effects of a monetary policy contraction under flexible exchange rates
Output, У
of aggregate demand begins a gradual lall in the price level towards the lower price target. Thus, thc interest rate will have to be kept high for some time, and, as you saw in thc previous subsection, this must lead in the first year to an overshooting appreciation of thc nominal exchange rate to EB—beyond the appreciation to E, that occurs in the medium run. With the small pricc fall and large exchange rate- rise in thc first year, thc real exchange rate also appreciates, which further reduces aggregate demand via net exports. This is captured in the relatively flat slope of the AD curve, which implies a larger short- run fall in output to YB in the open economy.
In the next few years, the /4S curvc slowly shilts down from /1S(, towards /4S| as the expectcd price level (alls. As the cxpectcd and actual pricc levels slowly approach thc lower target value, output slowly returns to the natural level. In turn, the real exchange rate gradually depreciates back to its original value, while thc nominal exchange rate ends up stronger than it was prior to the policy change. Thc interest rate will have been brought back down to i„ = i*
To summarise: Under flexible exchange rates, a monetary policy contraction leads to a higher interest rate in the short run, an overshooting nominal exchange rate, a stronger real exchange rate and thus a fall in output. /4s prices gradually converge down to the central bank's lower price target, the economy returns in the medium run to the natural level of output, the same real exchange rate but an appreciated nominal exchange rate.
Rational or adaptive expectations
Compare this with the ► conclusions we drew for the closed economy in Chapter 7; there, it was real investment that had to change to compensate for the real shock in the medium run.
Thus far, we have said little about how exchange rate expectations arc formed. I he qualitative results in Figure 21.8 follow lor most reasonable theories of expectations. The benchmark is rational expectations, where expectations turn out to be consistent with actual outcomes. Fven il they were adaptive, so that expectations adjust only gradually towards the actual appreciating path, we get a similar overshooting story for a monetary policy expansion. These overshooting exchange rate results after a monetary policy change were first discovered by the late Rudigcr Dornbusch in the mid-1970s. He showed how they occurred with rational or adaptive expectations.
A permanent fiscal shock
Now consider a permanent real demand shock, such as an increase in G. Any positive demand shock directly forces the IS and AD curves to shift to thc right in the short run. Since the natural level ol output is unchanged, net exports will have to fall to exactly compensate for the real shock in the medium run.
With an unchanged price level target, the nominal (and real I exchange rate must rise in the medium run if net exports are to tall. This causes the IS and AD curves to shitt back to their original positions. So, fiscal expansion leads to a real appreciation in the medium run.
II agents in financial markets rationally expect this medium-run exchange rate change, thc current exchange rate will change by this amount immediately, which will be exactly enough to cancel out the dircct aggregate demand effect. Therefore, with rational expectations, even in thc short run, thc IS and AD curves will shitt back to their original positions, and there will be no short-run or medium-run effects on output, unemployment and inllation from this real shock. Ol course, prior to the establishment of this short-run equilibrium (where NX falls by thc amount of thc fiscal expansion), there- may be J-curvc effects (which we discussed in Section 19.5), but these disappear fairly quickly. Wc now have a very powerful conclusion- with rational expectations, any real demand shocks, such as a fiscal shock, will have no effect on domestic economic activity, even in the short run. This fiscal policy ineffectiveness result under flexible exchange rates was first established by Robert Mundell in the 1960s.
The expected exchange ► rate in the second year will also be higher than the medium-run value E|, though less than Eg. This means that the short-run position of the interest parity curve must be IPS, which is flatter and to the right of the new medium-run curve./Pi.
Now what happens if exchange rate expectations are not set rationally? For example, they may adapt only slowly to what happens in the medium run. Then the nominal exchange rate will adjust less in thc short run than in the medium run, and so the IS and AD curves shift only partially back to their original positions—in the short run, thc cconomy can now experience changed output, prices and the domestic interest rate. With the gradual movement of the expected and thus thc actual cxchangc rate to thc medium-run value, there need be no overshooting after real shocks of this form.
21.4 CHOOSING BETWEEN EXCHANGE RATE REGIMES
Let us now return to the question that motivates this chapter. Should countries choose flexible exchange rates or fixed exchange rates? Arc there circumstances when flexible rates dominate, and others when fixed rates dominate?
Much of what we have seen in this and the previous chapter would seem to favour flexible exchange rates:
• Section 21.1 argued that the exchange rate regime may not matter in thc medium ain. But il is still the case that it does matter in the short ain. In the short ain. countries that operate under fixed exchange rates and perfect capital mobility give up two macroeconomic instruments, ihe interest rate and the exchange rate. This not only reduces their ability to respond to shocks but may also lead to exchange rate crises.
• Section 2 1.2 argued lhat the anticipation that a country that operates under a fixed exchange rate may have to devalue leads investors lo ask lor very high interest rales, making the economic situation worse and putting more pressure on the country to devalue—so, this is another argument against fixed exchange rates.
• Section 21.3 introduced one argument against flexible exchange rates—namely, that under flexible exchange rates the exchange rate may move a lot and may be difficult to control through monetary policy. In the short run, we can expect to see overshooting of the medium-ain exchange rate. To bring down thc price level or to disinflate), the economy must endure an overshooting appreciation ol ihe nominal and real exchange rates, which intensifies the necessary recession.
On net, it would appear that, Irom a macroeconomic viewpoint, flexible exchange rates dominate lixed exchange rates. This indeed appears to he the consensus that has emerged among economists and policy-makers.
The consensus goes like this: in general, flexible exchange rales are preferable, with two exceptions:
1. when a group of countries is already tightly integrated, in which case a common currency may be the right solution
2. when the central bank cannot be taisted to follow a responsible monetary policy under flexible exchange rates, in which case a strong iorm ol fixed exchange rates, such as a currency board or dollarisation, may provide a solution.
Let us discuss each ot these two exceptions.
Common currency areas
Countries that operate under a fixed exchange rate regime are constrained to have the same interest rale. But how costly is that constraint? II thc countries face roughly the same macroeconomic problems and the same shocks, they would have chosen similar policies in the first place. Forcing them to have the same monetary policy may not be much ol a constraint.
This argument was first explored hy Robert Mundell who looked at the conditions under which a i This is the same Mundell set of countries might warn to operate under fixed exchange rates or even adopt a common currency. who put together the For countries to constitute an optimal currency area Mundell argued, they need lo satisfy one of two Mundell-Fleming model conditions: ' Vой S3W in ChaPter 20-
• The countries have to experience similar shocks. We iust saw the rationale for this: if they have similar shocks, they would have chosen roughly the same monetary policy anyway.
• Or. if the countries experience dilterent shocks, they must have high lactor mobility. If workers, for example, are willing to move Irom countries that are doing poorly to countries that are doing well, lactor mobility rather than macroeconomic policy can allow countries to adjust to shocks. When the unemployment rate is high in a country, workers leave that country to take jobs elsewhere, and thc unemployment rate in that country decreases hack to normal. II the unemployment rate is low, workers come to the country, and the unemployment rate in the country increases back to normal. The exchange rate isn't needed.
Following Mundell's analysis, most economists believe, lor example, that the common currency area composed ol the six states and two territories of Australia is close to an optimal currency area. True, the first condition isnt satisfied: individual states and territories sullcr Irom different shocks. Western Australia is more affected by global shilts in demand lor minerals than Tasmania. New South Wales and Victoria are more affected by what happens to thc price of manufactured goods than South Australia, and so on But the second condition is largely satisfied. There is considerable labour mobility across Australia. When a state does poorly, workers leave that state. When it docs well, workers come to that state. State unemployment rates return gradually to normal not because of state-level macroeconomic policy but because of labour mobility.
And there are clearly many advantages to the use of a common currency. For lirms and consumers within Australia, thc benefits ol having a common currency arc obvious,- think ol how complicated lilc would he if you had lo change money every time you crossed a state border. I he benefits go beyond these lower transaction costs. When prices arc quoted in thc same currency, il becomes much easier for buyers to compare prices, and competition between firms increases benefiting consumers. Given these benefits and the limited macroeconomic costs, ii makes good sense for Australia to have a single currency.
FOCUS i BOX
In adopting the euro in 2002 twelve European countries made the choice lo move to a single currency. Look at the focus box Thc euro: A short history .) Is thc economic argument tor this new common currency area as compelling as it is lor Australia, or lor ihe United Slates?
THE EURO:A SHORT HISTORY
As the European Union (EU) celebrated its thirtieth birthday in 1988, several governments decided that the rime had come to plan a move to a common currency. They asked Jacques Delors. the president of the EU. to prepare a report, which he presented in June 1989. The Delors report suggested moving to a European Monetary Union (EMU) in three stages:
- Stage I was the abolition of capital controls.
- Stage II was the choice of fixed parities, to be maintained except in 'exceptional circumstances'.
- Stage III was the adoption of a single currency. Stage I was implemented in July 1990.
Stage II began in 1994, after the exchange rate crises of 1992-93 had subsided. A new institution, the European Monetary Institute (EMI), was created to work out both the details of the transition and the rules of the new regime. A minor but symbolic decision involved choosing the name of the new common currency.The French liked 'ecu' (European currency unit), which is also an old French currency name. But its partners preferred euro, and the name was adopted in 1995.
In parallel. EU countries held referendums on whether they should adopt the Maastricht Treaty. The treaty, negotiated in 1991, set three main conditions for joining the EMU: low inflation, a budget deficit below 3 per cent, and a public debt below 60 per cent. The treaty wasn't very popular, and in many countries the outcome of the popular vote was close. In France, the treaty passed with only 51 per cent of the votes. In Denmark, the treaty was rejected.
Each Australian state (or ► territory) could have its own currency that freely •loated against otner state currencies. But this isn't the way things are: Australia is a common currency area, with one currency, the Australian dollar.
In 1996-97, it looked as if few European countries would satisfy the Maastricht conditions. But several countries took drastic measures to reduce their budget deficit. When the time came to decide, in May 1998. which countries would be members of the euro area, eleven countries made the cut: Austria, Belgium. Finland, France, Germany, Italy, Ireland. Luxembourg, the Netherlands. Portugal and Spain. The United Kingdom. Denmark and Sweden decided to stay out. at least at the beginning. Greece didn't qualify. Stage III started in January 1999. Parities between the eleven currencies and the euro were 'irrevocably' fixed.The new European Central Bank (ECB). based in Frankfurt, became responsible for monetary policy for the euro area. In 2001.Greece finally qualified and joined.
Figure I Details of the euro banknotes
I *8r«?
Ш— •?
5 50 ШЬ !
ku'tttruM
Щ'Ш
i
100
...•10 100'"'°
10 im
!20 *Гггт
ft ш
200., 200
- V
201Ш
20»и» 20 200ЯИЬч
Г?
500 Ш.
Л)|
500
500wv,
SOURCE- European Central Bank.
From 1999 to 2002, the euro existed as a unit of account but euro coins and bank notes didn't exist. In effect, the euro area was still functioning as an area with fixed exchange rates.The next and final step was the introduction of euro coins and banknotes in January 2002. For the first few months of 2002, national currencies and the euro circulated side by side, after which national currencies were taken out of circulation.
Today, the euro is the only currency used in the 'euro area', as the group of sixteen member countries is called.
For more on the euro, go to
There is little question that a common currency will yield for Europe many ol the same benefits as it does lor large countries like the Llnited States. A report by the European Commission estimated that the elimination ol foreign-exchange transactions within the euro area led to a reduction in costs ol 0.5 per cent ol the combined CDP of these countries. There are also clear signs that the use of a common currency has increased competition. When shopping for cars, for example, European consumers are now looking lor the lowest euro price anywhere in the euro area. This has already led to a decline in the price of cars in several countries.
There is, however. less agreement on whether Europe constitutes an optimal common currency area. This is because neither of the two Mundell conditions appears to be satisfied. While the future may he different, European countries have experienced very different shocks in the past,- recall German reunification, and how differently it has affected Germany and the other European countries. And labour mobility is very low in Europe, and likely to remain so. Workers move much less within European countries than workers move within Ausiralia and the Llnited Slates. Given the language and cultural differences between European countries, mobility between countries is likely to be even lower. The risk, therefore, is that, at some time in the future, one or more euro members sullcrs Irom a large decline in demand and output, and can use neither the interest rale nor the exchange rate lo increase activity. As we saw in Section 21.1. the adjustment will still take place in the medium run. But, as you also saw there,
this adjustment may be long and painful. This is no longer a hypothetical worry: some euro countries, in particular Portugal sincc 2007, have been suffering from low output and a large trade deficit. Without the option of a devaluation, achieving a real depreciation may require many years of high unemploy¬ment and downward pressure on wages and prices in Portugal relative to thc rest ot thc curt) area.
An ANZ area?
Does it make sense (or Australia and New Zealand (and perhaps other local Pacific states) to adopt a single currency? To answer this, much research needs to be done to assess the costs and benefits for each party. This is a live issue that deserves carcful analysis. The issues confronting thc euro in Europe provide a template for this analysis. Labour mobility between Australia and New Zealand has been quite high, but the movement towards Australia has probably been disproportionately large. Clearly, the reasons aren't to cscapc higher unemployment in New Zealand.
What about trade between the two countries? For Australia, New Zealand is a minor partner—6 per cent of exports go to New Zealand, and 3 per cent of imports come from that country. For New Zealand. Australia is an important partner—12 per ccnt of exports go to Australia, and 32 per cent ol imports come from Australia. So, thc gains from reducing thc costs of trade will mostly benefit New Zealand, which suggests that the push for a common currency will come from there.
The two economies have many similarities—both are signilicantly altected by global commodity prices, though the proportion of rural (particularly dairy) commodities that matter in New Zealand is higher. Looking at both the nominal and real effective exchange rates, thc variance has been twice as large in New Zealand. Nevertheless, the patterns have been lairly similar over the last fifteen years, as Figure 2 1.9 shows. This suggests that the arguments tor a common currency won't go away.
Hard pegs, currency boards and dollarisation
The second case lor tixed exchange rates is very different Irom the lirst. It is based on the argument that there may be times when a country wants lo limit its ability to use monetary policy. We will look at this argument in more detail in Chapter 24, where we look at thc dynamics of hyperinflation, and in Chapter 26, where we look at monetary policy in general. But the essence of the argument is simple.
Look at a country that has had very high inflation in thc recent past. This may be, tor example, because it was unable to finance its budget deficit by any other means than through money creation, resulting in high money growth and high inllation. Suppose that the country decides to reduce money growth and inllation. One way ol convincing linancial markets that it is serious about reducing money
£ ш
Real Nominal
Figure 21.9 Nominal and real effective exchange rates in Australia and New Zealand
Though New Zealand's nominal and real exchange rates were more variable than Australia, they followed the same pattern. SOURCE IMF, International Financial Statistics.
EXO 1ANGE RATE REGIMES chaptcr 21
growth is to fix its exchange rate. Thc need to use the money supply to maintain the parity then ties the hands of the monetary authority. To the extent that linancial markets expect the parity to he maintained, they will stop worrying about money growth being used to finance the budget dcficit.
Note thc qualifier, to the extent that linancial markets expect the parity to be maintained. Fixing thc exchange rate isn't a magic solution. Thc country needs to convince financial investors that not only is the exchange rate fixed today but that it will remain fixed in thc future. This has two implications:
1. Fixing the exchange rate must be part ol a more general macroeconomic package. Fixing the exchange rate while continuing to run a large budget delicit will only convince financial markets that money growth will start again, and that a devaluation is soon to come.
2. Making it symbolically or technically harder to change the parity may also be useful, an approach known as a hard peg.
An extreme lorm of a hard peg is simply to replace thc domestic currency with a foreign currency. Because the foreign currency typically chosen is the US dollar, this is known as dollarisation.
Few countries are willing, however, to give up their currency and adopt the currency ol another country. A less extreme way is the use ol a currency board. Under a currency board, a ccntral bank stands ready to exchange foreign currency for domestic currency at the official exchange rate,- furthermore, it cannot engage in open-market operations—that is, buy or sell government bonds.
гас
Г П/Г-V
BOX
Perhaps the best-known example ol a currency board is that adopted by Argentina in 1991 but abandoned in a crisis at thc end ol 2001. Thc story is told in the focus box 'Argentina's currency board'. Economists differ on what conclusions one should draw Irom what happened in Argentina. Some conclude that currency boards aren't hard enough: they don't prevent exchange rate crises. So, il a country decides to adopt a lixed exchange rate, it should go all the way and dollarisc. Others conclude that fixed exchange rates arc a bad idea. If currency boards arc used at all, they should be used for a short period ol time, before the country returns to a lloating exchange rate regime.
ARGENTINA'S CURRENCY BOARD
When Carlos Menem became president of Argentina in 1989, he inherited an economic mess. Inflation was running at more than 30 per cent a month. Output growth was negative.
Menem and his economy minister, Domingo Cavallo. quickly came to the conclusion that, under the circumstances, die only way to bring money growth—and, by implication, inflation—under control, was to peg the peso (Argentina's currency) to the US dollar, and to do this through a very hard peg. So in 1991 Cavallo announced that Argentina would adopt a currency board. The central bank would stand ready to exchange pesos for US dollars, on demand. Furthermore, it would do so at the highly symbolic rate of one US dollar for one peso.
The creation of a currency board and the choice of a symbolic exchange rate had the same purpose: to convince financial markets that the government was serious about the peg, and to make it more difficult for future governments to give up the parity and devalue. Making the fixed exchange rate more credible in this way would, it was hoped, decrease the risk of a foreign-exchange crisis.
For a while the currency board appeared to work extremely well. Inflation, which had exceeded 2,300 per cent in 1990. was down to 4 per cent by 1994! This was clearly the result of the tight constraints the currency board put on money growth. Even more impressive, this large decrease in inflation was accompanied by strong output growth. Output growth averaged 5 per cent a year from 1991 to 1999.
Starting in 1999, however, growth turned negative, and Argentina went into a long and deep recession. Was the recession due to the currency board? Yes and no:
4 When Israel was suffering from high inflation in the 1980s. an Israeli finance minister proposed dollarisation as part of a stabilisation program. His proposal was perceived as an attack on the sovereignty of Israel, and he was quickly fired.
• Throughout the second half of the 1990s, the US dollar steadily appreciated against other major world currencies. Because the peso was pegged to the US dollar, the peso also appreciated. By the late 1990s. it was clear that the peso was overvalued, leading to a decrease in demand for goods from Argentina, a decline in output, and an increase in the trade deficit.
• The currency board wasn't fully responsible for the recession.There were other causes as well. But the currency board made it much harder to fight it. Lower interest rates and a depreciation of the peso would have helped the economy recover, but under the currency board this wasn't an option.
In 2001, the economic crisis turned into a financial crisis and an exchange rate crisis, along the lines described in Section 21.2:
• Because of the recession, the fiscal deficit increased, leading to an increase in government debt. Worried that the government might default on its debt, financial investors started asking for very high interest rates on government debt, making the fiscal deficit even larger and. by doing so, further increasing the risk of default.
• Worried that the government would give up the currency board and devalue in order to fight the recession, financial investors started asking for very high interest rates in pesos, making it more costly for the government to sustain the parity with the US dollar, and so making it more likely that the currency board would be abandoned.
In December 2001, the government defaulted on part of its debt. In early 2002, it gave up the currency board and let the peso float. By October 2002, the peso had lost half of its real effective value since January 2002. As a consequence, in 2002, import volumes shrunk by almost half, though exports had hardly improved eighteen months later. Real GDP fell II per cent in 2002, unemployment reached 20 per cent and inflation went down to - I per cent.
Why did GDP fall so much after the peso had suffered a 50 per cent real depreciation? Didn't we show in Chapter 19 that a depreciation will raise aggregate demand through exports? What is going on?The answer is that large depreciations may initially have negative effects on aggregate demand—economists call these perverse effects 'balance sheet effects', and they were also very much in evidence during the Asian financial crisis in 1997-98. Many firms and banks had borrowed short-term funds from international capital markets, which were denominated in US dollars.When the peso collapsed, the domestic value of these US-dollar debts increased substantially, and the debtors were unable to repay. The banks were in trouble and bank runs started. The banks were in no position to lend funds to firms, and depositors couldn't get their money out. This meant that consumption and investment, and thus aggregate demand, fell in the first year after the big depreciation. Exports could be expected to rise after a few years, and eventually more than cancel out the short-run negative effects, but until exports responded the outlook for Argentina was glum. Output began to slowly improve, but it took until 2005 for GDP to reach its 1998 level. Does this mean that the currency board was a bad idea? Economists still disagree.
• Some argue that it was a good idea but that it didn't go far enough. Argentina should simply have dollarised—that is, adopted the US dollar as the currency and eliminated the peso altogether. By eliminating the domestic currency, this solution would have eliminated the risk of a devaluation. The lesson, they argue, is that a currency board doesn't provide a sufficiently hard peg for the exchange rate. Only dollarisation will do.
• Others argue that the currency board may have been a good idea at the start but that it shouldn't have been kept for so long. Once inflation was under control, Argentina should have moved from a currency board and returned to a floating exchange rate regime.The problem is that Argentina kept the fixed parity with the US dollar too long, to the point where the peso was overvalued and an exchange rate crisis was inevitable.
The debate is likely to continue. Meanwhile. Argentina still has work to do to reconstruct its economy.
Even under a fixed exchange rate regime, countries can adjust their real exchange rate in thc medium run. They can do so by relying on adjustments in the price level. Nevertheless, the adjustment may he long and painful. Exchange rate adjustments allow thc cconomy to adjust faster, and thus reduce the pain that comes from a long adjustment.
Exchange rate crises typically start when participants in financial markets believe a currency may soon be devalued. Defending the parity then requires very high interest rates, with potentially large adverse macroeconomic effects. I hese adverse effects may lorce the country to devalue, even il there were no initial plans for such a devaluation.
The exchange rate today depends on both (11 thc difference between current and expected future domestic interest rales and currcnt and expcctcd luture foreign interest rates, and (2) the cxchangc rate expected in the future.
- Any factor that increases current or expected future domestic interest rates leads to an increase in the exchange rate today.
- Any factor that increases currcnt or expected future foreign interest rates leads to a decrease in thc exchange rate today.
- Any lactor that changes expectations of thc exchange rate in the future leads to a change in the exchange rate today (for example, luiure commodity prices in the case of Australia).
If interest rates change and remain so for some time, then the exchange rate will overshoot its medium-run value in ihe short run.
A contractionary monetary policy will reduce prices in the medium run, but the overshooting exchange rate will amplify the temporary recession beyond that occurring in a closcd economy. If expectations arc rational, a fiscal shock will have little clleci on output hut will affcct the real exchange rate and net exports.
There is wide agreement among economists that flexible exchange rate regimes generally dominate fixed exchange rate regimes, except in two cases;
I I he first is when a group of countries is highly integrated and lorms an optimal currency area.
You can think of a common currency tor a group of countries as an extreme form ot fixed exchange rates among this group of countries. For countries to form an optimal currency area, they must either facc largely similar shocks or there must be high labour mobility between these
countries.
optima! currency area, 499 Maastricht Treaty, 500 European Central Bank (ECB), 500 hard peg, 503 dollarisation, 503 currency board, 503
gold standard, 485 realignments 487 overnight interest rale, 488 lloat, 488
world commodity prices 140 overshooting, 193
2. The second is when a central bank cannot be trusted to follow a responsible monetary policy under flexible exchange rates. In this case, a strong lorm ol fixed exchange rates, such as dollarisation or a currency hoard, provides a way of tying thc hands of the central bank.
Quick check
1. Using the information in this chaptcr, label each of the following statements 'true', 'false' or
'uncertain'. Explain briefly.
a. Britain's return to the gold standard caused years of high unemployment.
h. A sudden fear that a country is going to devalue may force an exchange rate crisis even il the lear initially had no basis.
c. Because speculative behaviour by foreign investors can cause currency crises, small countries would be better olf not allowing foreigners to hold domestic assets.
d. High labour mobility within Europe makes the euro area a good candidate lor a common currency.
e. Since lixed exchange rale economies return eventually to their natural level of output, there is never a reason to devalue.
f. Changes in the expected level ol the exchange rate lar in the future have little effect on the current level ol thc exchange rate.
g. The Australian dollar today is likely to strengthen it there is an increase in expected future world commodity prices.
2. Consider a country operating under fixed exchange rates, with aggregate demand and aggregate
supply given by equations (21.1) and (21.2):
Y - C(Y - T) + l(Y,i - тг1') + G + NX
Assume lhat the economy is initially in medium-run equilibrium, with a constant price level and output equal to the natural level of output. Foreign output, the foreign price level and the foreign interest rate are fixed throughout the problem. Assume that expected (domestic) inflation remains constant throughout the problem.
a. Draw an AS-AD diagram (or this economy.
b. Now suppose that there is an increase in government spending. Show the eflects on the AS-AD diagram in the short шп and the medium run. How do output and thc pricc level change in the medium run?
c. What happens to consumption in the medium run?
d. What happens to thc real exchange rate in the medium run? (Hint: Consider thc effcct on the price level you identified in part (b).) What happens lo net exports in the medium run?
e. Civen that ihe exchange raic is fixed, what is lhe domestic nominal interest rate? Does the increase in government spending alfect the domestic nominal interest rate? What happens to the real interest rate in the medium run? (Hint: Remember that expected inflation remains constant by assumption.) What happens to investment in the medium am?
f. In a closed economy, how docs an increase in government spending affect investment in the medium run? (Relet to Chaptcr 7 il you need a refresher, i
g. Comment on the following statement. In a closed economy, government spending crowds out investment. In an open economy with fixed exchange rates, government spending crowds out net exports.'
3. Nominal and real interest parity
In equation (18.4), we wrote the nominal interest parity condition as
E,
In the appendix to this chapter, we derive a real interest parity condition. We can write the real interest parity condition in a manner analogous to equation (18.4):
r, » r, -
a. Interpret this equation. Under what circumstances will thc domestic real interest rate exceed the foreign real interest rate?
Assume that the one-year nominal interest rate is 10 per cent in the domestic economy and 6 per cent in the foreign economy. Also assume that inflation over the coming year is expected to be 6 per cent in the domestic economy and 3 per cent in the foreign economy. Suppose that interest parity holds.
b. What is the cxpccied nominal depreciation of the domestic currency over the coming year?
c. What is the expected real depreciation over the coming year?
d. Il you expected a nominal appreciation ol the currency over the coming year, should you hold domestic bonds or foreign bonds?
Devaluation and interest rates
Consider an open economy with a fixed exchange rate, E. Throughout the problem, assume that the foreign interest rate, i*, remains constant.
a. Suppose that linancial market participants believe that the government is committed to a fixed exchange rate. What is the expected exchange rate? According to the interest parity condition, what is the domestic interest rate?
b. Suppose that financial market participants do not believe that the government is committed to a fixed exchange rate. Instead, they suspect that the government will cither devalue or abandon the lixed exchange rale altogether and adopt a flexible exchange rate. If the government adopts a flexible exchange rate, financial market participants expect the exchange rate to depreciate from iis currciu lixed value, £. Under these circumstances, how docs tjhc cxpectcd exchange rate compare with E? I low does the domestic interest rate compare with E?
c. Suppose that linancial market participants leaied a devaluation, as in pari (b), and a devaluation actually occurs. The government announces that it will majntain a lixed exchange rate regime but changes ihe level ol thc fixed exchange rate toE' whereE' < E. Suppose that linancial market participants believe that thc government will remain committed to the new exchange rate, E', and that there will be no further devaluations. What happens lo the domestic interest rate after the devaluation?
d. Does a devaluation necessarily lead lo higher domestic interest rates? Docs fear of a devaluation necessarily lead to higher domestic interest rates?
Dig deeper
S. Self-fulfilling exchange rate crises
Consider an open economy with a fixed exchange rate, L. Suppose that, initially, financial market participants believe that the government is committed to the fixed exchange rate. Suddenly, however, financial market participants become fearful that the government will devalue or allow the exchange rate to float (a decision that everyone believes will cause the currency to depreciate).
a. What happens to thc expected exchange rate, EJ',,? (Sec your answer to problem 4(b).) Suppose that, despite the change in the expected exchange rate, the government keeps the exchange rate fixed today. Let II' stand for (uncoveredI interest parity.
b. Draw an IS-LM-IP diagram. How does the change in the expcctcd exchange rate affect the IP curve? As a result, how must the domestic interest rale change to maintain an exchange rate of E?
c. Given your answer to pari (b), what happens to thc domestic money supply if the central bank defends the fixed exchange rate? How docs thc LM curve shift?
d What happens to domestic output and the domestic interest rate? Is it possible that a government that was previously committed to a lixed exchange rate might abandon it when faced with a lear ol depreciation (either through devaluation or through abandonment of the fixed exchange rate regime)? Is it possible that unlounded tears about a depreciation can create a crisis? Explain your answers.
The macroeconomics of flexible exchange rates
Using the model underlying Figure 21.5, determine the short-run effects, dynamics and medium-rim effects of:
An increase in government expenditure, G.
An increase in thc foreign interest rate, i*.
An increase in the markup lactor /j..
The Australian dollar is positively affected by world commodity prices. So, assume now that net exports also depend on an exogenous parameter y, which represents the positive effect of \world commodity prices on the real value of exports—an increase in у leads to higher exports. The net export function becomes NX/e,Y,Y",y/. Using this in the model underlying Figure 21.5. explain why the exchange rate appreciates in the short run and medium run in response to
An increase in world commodity prices, y. Exchange rate overshooting
a. Suppose that there is a permanent easing of monetary policy that leads to a 10 per cent increase in the price level target in a closed economy. What is the effect on actual price and the money supply in the medium run? Hint: II you need a refresher, review the analysis in Chapter 7. i
In the closed economy, we said that money was neutral because in the medium run a change in the money stock affected only the price level, with no effects on any real variables. In theory, money is also neutral in the open economy. In the medium run, a change in the price level target and thus the money stock will not affect lhe real exchange rale, although il will affect lhe price level and the nominal exchange rate.
b. Consider an open economy with a flexible exchange rate. Write down the expression lor thc real exchange rate. Suppose that there is a 10 per cent increase in the price level target and assume that it has the same effcct on the money stock in the medium run as you lound in part (al. II the real exchange rate and the foreign pricc level are unchanged in thc medium run, what must happen to the nominal exchange rate in the medium run?
c. Suppose that it takes n years to reach the medium run (and everyone knows this). Given your answer to part b), what happens to E;'_„ (the expected exchange rale lor the time n periods Irom now) alter a 10 per cent increase in the price level target?
d. Consider equation (21.7). Assume that the loreign interest rate is unchanged lor the next n periods. Also assume for the moment, that the domestic interest rate is unchanged for the next n periods. Given your answer to part (c), what happens to the exchange rate today (at time t) when the policy change occurs?
e. Now assume more realistically that the new policy causes the domestic interest rate to fall now and between time I and time t + n. Again assume that the foreign interest rate is unchanged. Compared with your answer to part (d). what happens to the exchange rate today 'at time Г? Does thc exchange rate move more in the short run than in the medium run?
The answer to part lei is yes. In this case, the short-run depreciation is greater than the medium-run depreciation. This phenomenon is called overshooting, and may help to explain why the exchange rate is so variable.
Explore further
8. Exchange rates ami expectations
Lt-1
a.
b.
In this chapter we emphasised that expectations hare an important effect on the exchange rate. In this problem we use data to gel a sense of how large a role expectations play. Using the results in Appendix 2 at the end of the book, you can show that the uncovered interest parity condition, equation (21.6), can be written as
• ф - (ii-j
In words: The percentage change in the exchange rate (the appreciation of the domestic currencyI is approximately equal to the change in the interest rate differential (between domestic and foreign interest ratesI plus the percentage change in exchange rate expectations (the appreciation of the expected domestic currency value). IVe will call the interest rate differential the 'spread'.
Go to the website ol the Bank ot Canada (www.bankbanquc-canada.ca) and obtain data on thc monthly one-year Treasury bill rale tor the past ten years. Download the data into a spreadsheet. Now go to thc website ot the federal Reserve Bank ol Si I.ouis (http://rcscarch.silouisled.org Ired2 i and download data on the monthly LIS one-year Canadian Treasury bill rale lor the same lime period. (You may need lo look under 'Constant Maturity Treasury securities rather than Treasury Bills'.) For each month, subtract the Canadian interest raie from the LIS interest rate to calculate ihe spread. Then, lor each month, calculate the change in thc spread from the preceding month. (Make sure to convert the interest rate data into the proper decimal form.)
Ai the website of the St Louis Fed, obtain data on the monthly exchange rate between thc LIS dollar and ihe Canadian dollar tor ihe same period as your data Irom pari (a Again, download the data into a spreadsheet. Calculate thc percentage appreciation of the LIS dollar for each month. Using the standard deviation function in your software, calculate the standard deviation ol thc monthly appreciation of thc LIS dollar. The standard deviation is a measure ol the variability of a daia series.
c. For each month, subtract the change in ihe spread (part [a]) from the percentage appreciation of the dollar (part [b]). Call this difference the change in expectations. Calculate the standard deviation of thc change in expectations. 1 low docs il compare with the standard deviation of the monthly appreciation of the dollar? There are some complications we do not take into account here. Our interest parity condition does not include a variable that measures relative asset demand. We explored the implications of changes in relative asset demands in Appendix I of Chapter 20. In addition, changes in interest rates and expectations may be related. Still, the gist of this analysis survives in more sophisticated work. In the short run, observable economic fundamentals do not account for much of the change in the exchange rate. Much of the difference must be attributed to changing expectations.
We invite you to visit the Blanchard-Sheen page on the Pearson Australia website at
www.pearson.com.au/highered/blanchardsheen3e
for many World Wide Web exercises relating to issues similar to those in this chapter.
APPENDIX: THE REAL EXCHANGE RATE, AND DOMESTIC AND FOREIGN REAL INTEREST RATES
We derived in Section 21.3 a relation between the current nominal exchange rate, current and expected future domestic and foreign nominal interest rates, and the expected future nominal exchange rate (equation [21,7]).This appendix derives a similar relation, but in terms of real interest rates and the real exchange rate. It then briefly discusses how this alternative relation can be used to think about movements in the real exchange rate.
Deriving the real interest parity condition
Start from the nominal interest parity condition, equation (18.2):
1 + /, = £, (1 + О ТГ-
Recall the definition of the real interest rate from Chapter 14, equation (14.3):
, . _ (1
1" TT^fT
where тгj = (Pj+1 - Pt)IP, is the expected rate of inflation. Similarly, the foreign real interest rate is given by
two
(1 + 0
1 + r,:
(1 +
where 7r',c = (P^ti - P])IP, is the expected foreign rate of inflation.
Use these two relations to eliminate nominal interest rates in the interest-parity condition, so.
£,(1 + ~'tc)
1 +'."<1 <2IAI>
Note from the definition of inflation that (1 + тт?) = Pf^i/Pt and, similarly, (1 + тт',') = P',l\IP',. Using these relations in the term in brackets gives
E,d + -4 = £,p;cM/P; £5+1 (1 + *f) £,%iPf+,/Pt
Reorganising terms:
£^,/p; = wr,
Using the definition of the real exchange rate at time t and time t + 1:
£ftlf*0/P?+ 1
1 + = (1 + r't) ~
^t+1
(1 + rj
Replacing in equation (21 A. I) gives
Or. equivalently.
(2IA-2>
The real exchange rate today depends on the domestic and foreign real interest rates this year, and the expected future real exchange rate next year.This equation corresponds to equation (21.6) in the text, but now in terms of the real rather than the nominal exchange rate and interest rates.
Solving the real interest parity condition forward
The next step is to solve equation (2IA.2) forward, exactly in the same way as we did for equation (21.6) in the text.The equation above implies that the real exchange rate in year t + 1 is given by
_ (1
€'t1 " (1 + r*+1)
Taking expectations, as of year t:
_ (1 + г;.,) - (1 + rfc,)
Replacing in the previous relation:
(1 + r,)(1 + ft*i) - 1 (1+0(1+^,)
Solving for and so on gives
(1 + 0(1 + fa)... (1+ ru„)
(1 + rj)(1 + r't%i) ... (1 + rX)
This relation gives the current real exchange rate as a function of the difference between current and expected future domestic real interest rates and current and expected future foreign real interest rates, and of the expected real exchange rate in year t + n.
The advantage of this relation over the relation we derived in the text between the nominal exchange rate and nominal interest rates, equation (21.7). is that it is typically easier to predict the future real exchange rate than to predict the future nominal exchange rate. If. for example, the economy suffers from a large trade deficit, we may be fairly confident that there will have to be a real depreciation—that e%n will have to be lower. Whether there will be a nominal depreciation—what happens to £;,.„—is harder to tell. It depends on what happens to inflation, both at home and abroad over the next n years.
Pathologies
Sometimes (macroeconomic) things go very wrong. There may be a sharp drop in share prices, or a dramatic rise in the cost of borrowing. Or output falls significantly. Or unemployment remains high
for a long time. Or inflation increases to very high levels. These pathologies are the focus of the next three chapters.
CHAPTER 22
Chapter 22 looks at the macroeconomic analysis of financial crises. It looks at why confidence may deteriorate in financial markets, leading to a collapse in stock markets, insolvency in financial institutions, a cessation of bank lending, a rise in credit interest rates, and a decline in output. The example used in this chapter is the 2008 global financial crisis that originated in the sub-prime mortgage defaults in the United States.
CHAPTER 23
Chapter 22 focused on the short run, but sometimes these crises have longer run effects. Chapter 23 looks at depressions and slumps, periods during which output drops far below, and stays far below, its natural level. The chapter discusses the adverse effects of deflation, and what happens when an economy is caught in a liquidity trap. It then looks at the Great Depression, what triggered it, what made it so bad, and what eventually led to a recovery. It then turns to the current Japanese economic slump, a slump that started in the early 1990s and is still ongoing today. It shows that many of the factors that contributed to the Great Depression are also operating in Japan today.
CHAPTER 24
Chapter 24 looks at episodes of high inflation, from Germany in the early 1920s to Latin America in the 1980s and early 1990s, and Zimbabwe now. It shows the role of both fiscal policy and monetary policy in generating high inflation. Budget deficits lead to high nominal money growth. High nominal money growth leads to high inflation. Chapter 24 then looks at how high inflations end, and at the role and nature of stabilisation programs.
CHAPTER ^
The Macroeconomics of Financial Market Crises
F
inancial crises are nothing new. They occur with alarming, though unpredictable, regularity. Over the last two centuries, a financial crisis has happened on average every ten years. In this chapter we study fragile financial systems, seeing how crises might emerge and understanding how we might use macroeconomic theory to analyse the real effects. Of course, crises are not all of the same severity, and they range from those with only localised effects to those that lead to a long-lasting global macroeconomic depression. In Chapter 23 we study the Great Depression of the 1930s in detail, showing how it was initially associated with a financial crisis. Here we unravel the anatomy of a financial crisis, using the 2008 global financial crisis as an example.
• Section 22.1 shows how we can think of banks as financial intermediaries that bridge the gap between saving and investment, or between savers and borrowers, and it shows what happens to a bark's balance sheet when some of its assets go bad.
• Section 22.2 turns to some of the macroeconomic models we have developed in the book, extending them to the case
• Section 22.3 looks at the global financial crisis of 2008 that I explaining its roots and how it evolved and spread into othe
SOURCt The Economist 1987.
of a financial crisis.
segan with the sub-prime mortgage crisis in the United States, r financial markets and to other countries.
22.1 FINANCIAL INSTITUTIONS AS INTERMEDIARIES IN CRISIS
In our first look at financial markets in The Core (Chapter 4), we focused on banks as financial institutions whose main function was to issue money. We explained that banks accepted deposits and invested these in loans, bonds and reserves. To keep things simple for our story' about thc money supply process, we didn't worry about distinguishing loans Irom bonds. However, in thc focus box 'Hank mns' wc recognised that the distinction could be very important because it helped to explain why the financial system seems so Iragile Irom time to time. In this chapter we want to make the distinction because we want to emphasise the role ol financial institutions as intermediaries between savers and borrowers. We will introduce a richer menu of bonds to include securitised bonds.
11 It MACROECONOMICS OF l INANCIAL MARKET CRISL5
Table 22.1 ABC Bank's balance sheet
Assets Liabilities
Loans: $100 Deposits: $200
Bonds: $200 Debt: $50
Bank capital: $50
chapter 22
A primary function of hanks is financial intermediation—to match thc saving flows in an economy with the real investment llows. These investments are always risky, and the financial system has evolved to spread the risks across a wide range of savers. The innovations in finance over the last thirty years have been truly remarkable in finding ways to spread these risks widely. Unfortunately, evolution is never a one-way street towards perfection. I here arc many casualties as the survival ol thc fittest principle works its way out. And sometimes the fallout is system-wide. When the system goes into crisis, intervention becomes essential. Let's see how a crisis can occur in thc financial system.
Wc begin with a simplilicd balancc sheet of a typical bank. ABC Bank, which is involved in financial intermediation. This is shown in Table 22.1. A good joke going
around the Internet: 'There are two sides to bank balance sheets: the left and the rignt (liabilities and assets, respectively). On the left side there is nothing right, and on the right side there is nothing left.' (It's true we have a different left and right in Table 22.1. But it's still a
ABC Bank has taken deposits of $200 (millions, il you like, which arc a liability. It has also taken on good joke') other liabilities, which we will label debt, or borrowings Irom other banks, to the value ol $50. It has fulfilled its function as a financial intermediary by channelling $100 ol these deposits and debt to firms or households as loans, and $200 to other banks, lirms or the government in the lorm ol bonds including a special lorm ol bond, which we will explain shortly). Its assets are therefore worth $300, leaving $50 as bank capital owned by the shareholders ol ABC Bank but held as back-up lunding.
Both its loans and its bonds are risky, but lor different reasons. Its loans are made directly to ABC's own customers. ABC. expects them to be able to service or repay their loan in thc luture, but there is always a risk. However. ABC knows these customers and has a good idea of the risks involved, and optimally keeps some bank capital as a precaution. The bonds arc different. Apart Irom liquid assets, they include bonds as securitised loans issued by other financial institutions. These other institutions buy up the direct loans (naturally including sub-prime mortgages) ol various banks, package them up or securitise them) into 'bonds' and sell them on lo a bank like ABC. Of course. ABC has no relationship with the original borrowers (including sub-prime ones whose loans arc included in the securitised bonds. Further, thc securitised bonds are financially engineered to be very complex instruments. Therefore, ABC has a hard job computing the actual risks involved in the bond, and unfortunately allocates too little precautionary bank capital, i liven risk rating agencies, like Moody's, experience great diificuliy in evaluating the risks ol ABC's securitised bond assets, and continue ю give thc bank a sale credit rating,- it gives ihe bank an Aaa rating, which indicates that ihe credit risk is virtually zero.) All is well, so long as thc original loans in the securitised bonds perform well.
Unfortunately lor ABC, house priccs start falling and thc equity that home-owners have in their home declines. When that equity approaches zero, thc owners begin to foreclose on their mortgages. 1 bus many ol ihe mortgage loans in ABC's sccuriiiscd bonds go into default. Suppose thai one-quarter of ABC's bond holdings arc lost, so that they arc now worth only $150. ABC's direct loans of $100 arc siill fine because it selected customers with good credit ratings. ABC also knows from thc media that there are other financial institutions facing similar difficulties.
There are two issues that ABC i in common with many other banks) has to lace: bank capital shortage and bank runs
• Capital shortage. ABC now has assets worth only ($100 +■ $150) $250 which now equals its liabilities. Therefore its bank capital is zero It has no bank capital as a precaution against deposit withdrawals or against subsequent loan or bond failure. ABC is on the edge ol insolvency. It needs lo raise capital. How can it do this?
- It could borrow more money from other hanks, hut unfortunately many other banks are experiencing the same problem, liach bank knows precisely its own balance sheet difficulties, but it doesn't know what is happening with other banks. These banks arc becoming less willing to tie up their balance sheets in assets that cannot be easily sold in the current climate. Thus, borrowing from banks becomes available lor only short horizons and at veiy high interbank interest rates. Worse, it is possible that the debt ABC has borrowed already may be demanded back by other banks that are experiencing balance sheet difficulties.
- It could try to raise more money Irom shareholders to add to its liquid assets, but its share price- has already fallen a lot due to its financial troubles.
- II ABC cannot get more capital, it could sell some of its remaining assets. But these assets cannot be sold right away, unless ABC' is prepared to sell them at very low prices, in which case its bank capital and share price would collapse even further.
- What ABC won't do is increase its lending. Anyone coming to ABC asking for new credit will be denied.
Вап'к runs. The savers who deposited their lunds at ABC.! arc likely to know that ABC is in trouble, and so will want to withdraw their money. But ABC will not have enough spare liquid assets to return depositors' money. It will have to keep its doors locked, switch oil its ATMs and disconnect its Internet banking.
ABC may have to close, or it may be taken over by a bigger and healthier bank. Imagine this ABC. hank experience happening over thc whole financial system, and this is very much what happened in September and October 2008 in the Llnited States and then Europe. Assets associated with sub-prime mortgages went bad due to low housing prices that led to many foreclosures. Many financial institutions held securitised bonds containing these mortgage assets and so suffered severe balance sheet losses which led all institutions to worry about thc solvency ol other institutions. Normally banks lend and borrow between themselves on a daily basis in the interbank money market at the interbank interest rate. They stopped normal lending to each other, and so interbank interest rates went through the roof (see the focus box An indicator of the credit crunch—the TED spread'). Normal lending to households and lirms either ceased or was available for only short terms and at much higher interest rates. This is what is meant by the credit crunch, or sometimes the credit freeze.
Policy-makers did respond to this crisis. Policies were put in place to help banks whose balance sheet was in trouble. Central banks advanced kinds to banks to try to improve their capital positions. Public capital was put into major financial institutions directly associated with the housing finance problem Fannie Mac, Freddie Mac and also into many other privately owned banks. Further, many governments extended deposit guarantees to protect institutions Irom the danger ol bank runs, which occur when depositors panic and withdraw their lunds. The guarantees meant lhat any deposit in a bank eligible for a guarantee would be redeemed by the government in the event of a bank failure. As you would expect, institutions not eligible for a guarantee faced lhe prospect of depositors making rapid withdrawals and transferring their funds to guaranteed banks.)
However, ii is obvious at the time ol writing that not enough has been (or could be) done to avoid damage to the real economy. In late 2008, all countries were suffering Irom a rapid slowdown or decline in economic activity that was certainly attributable- to the global linanciai crisis. So how might we analyse the real macroeconomic ellects ol such a financial crisis on the real economy? We show you this in the next section.
This could happen to any of us: 'I went to the ATM this morning and it said "insufficient funds". I'm wondering, is it them or is it me?' ►
Go back and read again ► the focus box Bank
22.2 APPLYING MACROECONOMICTHEORYTO FINANCIAL CRISES
THl MACROECONOMICS OF FINANOAI MARKFT CRISES
chaptei 22
In this chaptcr wc will consider only the sliort-run implications ol a financial crisis leaving an analysis of longer lasting crises to thc next chaptcr In thc short rim expectations ol the luture output, interesi rates and liscal policy are likely to be very important.
• We made a distinction between the real and the nominal rate of interest с and i . the difference being the expected inflation rate see Chapter 14 lor more details).
Wc wont worry about inflation now. although il the 2008 crisis ends up in a depression (which seems unlikely), thc issue of deflation will be ot critical importance. (You will see more about depressions and deflation in the next chapter. Thcrelore. in this chapter we will assume constant expected inflation, so that real and nominal interest rates move together. For convenience, in thc rest of this section wc will talk only about thc nominal interest rate. i.
The market interest rate, adjusted for risk
The interest rate that we have studied up to now in the IS -LM model was the rate that thc central bank controlled—lor example, the cash rate in Australia or the federal funds rate in thc United States. This can be considered a risk-free rate However, most interbank lending and lending to lirms occurs at a higher market rate, the difference reflecting the risk of default.
Let's define thc credit default risk premium between a representative market interest rate and the risk-lree cash rate as .v. ' You can think ot .v as the TF.I) spread in thc focus box 'An indicator of the credit crunch—the TED spread. This means that thc market rate ol interest that firms pay to finance their real investments in plant, machinery, equipment and buildings is not i. but i i r. Let's call this thc market interest /у. This is given by:
i'M = / + x (22.1)
When the linancial sector is in good health, v is small. In a major financial crisis, x becomes very large. Indeed, tor some borrowers it may even become infinite—it bccomcs impossible lor them lo borrow.
Note ihat the IS curve, through real investment demand, depends on i^, not on i. Thc LAI curvc continues to depend on i because the appropriate opportunity cost ot holding money remains the interest rate on safe government bonds.
In Figure 22.1, we draw our IS and LM curves with i not iy) on the vertical axis and Y on the horizontal axis. Thc location of the IS curve depends on expectations ot future variables: output Y'*'), thc crisis risk premium (.r'1') and policy variables (/''', С', Vе). It ,v' rises and/or Y'1' falls, thc IS will shilt to the left. If future policy is expected to respond so thatfalls, and/or (tv rises and'or Vе tails the IS curvc will shilt to the right. Wc will explain these shilts as we go along.
We assume that the initial equilibrium is at Д, with output at its medium-run value, Y\, and the interest rate, /,„ determined by thc ccntral bank. As usual, with our intcrcst-ratc-sctting ccntral bank, thc LM curve will shitt endogenously.
Now consider what happens il .v increases from v(1 to v,, as in a financial crisis. This raises thc cost ol firms borrowing, which reduces their currcnt real investment spending, and thus lowers aggregate demand, forcing a leftward shift in the IS curve. In Figure 22.1. we represent this leftward shift from ISn to the curve labelled !St(i + .v, .
When a financial crisis occurs, it happens very last, with share prices collapsing and credit markets freezing up. In thc midst of the crisis, monetary policy can respond with the necessary speed. Since liscal policy requires much more time to execute, we will discuss its implications later in terms ot expected future changes in government expenditure and taxation.
Thc central bank may be able to control i with its monetary policy, but it cannot control .v. In the face of the crisis, the central bank has three broad choices-.
• Maintain the interest rate. It could decide to keep the policy interest rate at i(, in which case output would tall precipitously to Y,. This would actually require the ccntral bank to decrease the money supply from the market to maintain the original interest rate in the lace ol lower goods demand. The LM curvc would endogenously shift up to pass through Л, in Figure 22.1.
A financial crisis shifts the IS curve to the left. Depending on whether and how much the central bank responds, output will fall
Maintain the money supply, instead, it could choose to keep money supply constant, lowering the policy interest rate to i2. which would mean that the new equilibrium would be at on the original LM curve. Output would lall by less than belore, to reach
Expand the money supply. The central bank may decide that allowing such a large tall in output was undesirable in the context ol a serious linanciai crisis, and so it would Hood the market with money, pushing the interest rate down to /, in Figure 22.1, restoring output to its natural level. This is thc optimal response ol the central bank to this credit crisis shock—reduce the official interest rate by thc exact same amount as the credit risk default premium, v, rises.
There is a possible limit to this approach, and this is shown in Figure 22.2. Suppose that the new ISt curve shilts so lar lo the lelt on account of the crisis that il intercepts the horizontal axis al a level ot output lo the lelt of V,,. This possible limitation on monetary policy is known as the liquidity trap, and :i is an outcome that occurs when a financial crisis is so bad that it leads to a long-lasting depression in output. We will discuss this in much more detail in the next chapter. Here we jusi note that, once the interest rale reaches zero at A$, further expansion of money supply will not be able to increase output. This is because the interest rate cannot lall below 0
Now let's discuss the implications for current output of changing expectations ot future variables. An expected persistent crisis thigh x'' l. The current IS curve will shift even further to the left than shown in Figures 22.1 or 22.2 il the luture value ot x is expected to remain high. This is because the credit Ireeze is expected to persist and so future real investment will be restricted. Through the mechanisms discussed in Chapters 16 and 17, this will reduce currcnt spending. Current output will be lower and the task of current monetary policy more dillicult.
When i reaches 0. normal monetary policy has no effect. However. (
it is possible for the central bank to attempt quantitative easing, which is open market purchases of riskier assets than government bonds to try to reduce x.The Fed and the Bank of England tried this in 2008-09.
Output, У
"ATVIOI.OGIFS chapter 22
Figure 22.1
Financial crises in the IS-LM model
An expected fall in future output (low Y4'). The current IS curve will also shift further left than in Figures 22.1 or 22.2 ii ihe people expect luture output lo be much lower.
THF MACROECONOMICS OF FINANCIAL MARKET CRISLS
chapter 22
Figure 22.2
IS0(i + x0)
Financial crises and the zero interest rate limit
Yn
Output, У
It is possible that the IS curve shifts so far that the central bank cannot restore output to its original level.
In times ol crisis, there is considerable fear about the possibility ol a serious tall in economic activity. "T his is what is meant by a drop in confidence ol consumers and firms. Unlortunatcly. there- are many amateur predictors ol gloom and doom who make a name lor themselves by feeding on peoples lears. The most reliable forecasts at these times come Irom professional forecasters, such as those at central banks, the OECD and the IMF.
• An expected fall in the future policy interest rate (low i'eJ. II the central bank is expected to keep the policy interest rate low well into the future, the current IS curve will not shift as far left as shown in Figures 22.1 or 22.2. If low interest rates are expected to persist for some time real investment will fall by less.
However, it is not obvious that central banks will keep interest rates low for a long time. Why is thatr Because they will always be concerned that the necessary loose monetary policy during the crisis may sow the seeds ol higher luture inflation. Central banks will need to find the right balance between low interest rates in thc heat ot the crisis and rising interest rales in thc luture when the crisis is over.
• An expected expansion in future fiscal policy (high G'° and/or low T'CJ. The current IS curve will shift less to the lelt in Figures 22.1 or 22.2 it the government is expected to increase its fiscal deficit by raising spending and/or cutting taxes.
IS,О * x,)
In a financial crisis, future fiscal policy changes require careful thought. It is not a matter ot building more bridges to nowhere, or cutting taxes across the board. Governments should try to use tiscal policy surgically to address the key problems while stimulating weak demand. The govern-ment can choose to buy up the non-performing assets in financial institutions, but it has to lace up to the difficult question of at what price: thc current low market valuation, or a more realistic valuation. Alternatively, it can put public capital directly into the troubled banks, but it has to lace the problem of whether it wants to or should I take control of these banks It can guarantee deposits against a bank run for some time into the future, but lor how long, and for which banks?
As with monetary policy, governments shouldn i plan to increase spending or decrease taxes over the longer term in response lo a linancial crisis. Permanent values ol liscal policy variables should be constructed for normal times and not as a response to pathological crises. Some crises are all over within a year, with minimal damage lo the real economy, for these crises, only temporary liscal measures may be needed. Occasionally, crises are much deeper and last lor many years, in which case drastic and longer-lasting liscal policy changes may be necessary. The Greal Depression ol the 1930s is an example ol a deep crisis, which we take up in the next chapter.
Watch two comedians ► explaining the sub-prime mortgage crisis on youtube:
It is never easy tor policy-makers to know how and how much to respond in a crisis. While all crises have their key differences in detail our macroeconomic models provide an organising framework that helps us understand thc similarities. We now turn lo a discussion ot the key details ol the most recent linancial crisis.
22.3 THE GLOBAL FINANCIAL CRISIS OF 2008
Global linancial markets went into an acute crisis in September 2008, with stock-market prices tailing and bank lending grinding towards a halt. Global financial markets descended into a panic-selling mode as venerable LIS financial institutions got into serious financial trouble. Prominent examples were investment banks such as Bear Stearns Lehman Brothers and Merrill-Lynch, asset insurers like A1G (American International Group I, and secondary mortgage institutions like Fannie Mac Federal National Mortgage Association, set up in 19.38 and Freddie Mac (Federal Home Loan Mortgage Corporation, set up in 1970 to promote competition with Fannie Mac). Thc problems spread to Europe, with a number ot large institutions experiencing heavy losses on their balance sheets.
If you want to ► understand more about CDOs. CDSs or derivatives in general, search for therr in
The risk ol bank insolvency rose and banks became nervous about making transactions with other banks, in case these banks became insolvent. Banks no longer trusted each other like they do in normal times. I his trust is the key to the functioning of credit markets, and when trust fades credit markets do not operate normally. With stock-market prices falling and credit markets not functioning well, it became widely expected that economic activity would slow down, maybe leading to a global recession and in the worst-case scenario a depression. Central banks and government treasuries around thc world actively intervened in 2008 to try to avert a recurrence ol the Great Depression of the 1930s. At the time of writing (October 2008i. a recession, at thc least, seems inevitable.
The cause of the crisis
What was the primary cause ol this crisis: At its root was the biggest housing and credit bubble ever. This was mainly locatcd in the United States, where we have seen serious market failures in the provision of mortgage finance coupled with inadequate regulation ol the industry. There, when someone takes out a mortgage, the originating primary bank typically does not retain it on its balance sheet. Rather, it sells it usually lo government-sponsored secondary mortgage institutions like Fannie Mae and Freddie Mac. These secondary institutions either hold the mortgage or sell it again in highly complex securitised packages (which were called collateralised debt obligations, or CDOs, in thc jargon of financial markets) to investment banks and their investors. In addition, many of these financial institutions sold insurance contracts against possible default - using, in the iargon. derivative instruments callcd credit default swap, or CDSs) to holders ol the securitised bonds as well as to speculators. These insurance obligations were an additional item on the liability side of the balance sheets of these banks. Most ol this practice is fine, because it helps to spread risk to those that can and want to take it on. so long as regulations ensure that enough bank capital is held to cover thc risk ot the assets going in:o default and thc resultant increase in insurance obligations on thc liabilities side. But it can all go badly wrong if regulators do not regulate these practiccs and asset prices start falling.
Problems built up because the originating mortgage banks did not need to be too conccrncd about the risk ot their home-owner clients being unable to service their mortgage. Alter all, they had sold the mortgages to other financial institutions. As a result, they aggressively marketed and sold sub-prime mortgages lo many households, many ol whom were bound to default sooner or later, except in the unlikely event ot an ever-inflating housing price bubble. In many cases the mortgage originator got a
higher commission Irom seiling a sub-prime mortgage than Irom selling a prime one. These sub-prime mortgages had a very low honeymoon' interest rale for the first year or two, but after that the interest rate would jump well above the prime mortgage rate, a prohibitive rate lor many poorer lamilies. From 2008 many ol these 'out-of-honeymoon sub-prime mortgages started to go into default, and the banks were not prepared to refinance them at better rates, in part because housing prices had begun to lall, thus eliminating some ol the households' equity in their house. In lact, across the United States house prices fell almost 20 per cent in 2008, and many people expected another 10 per cent lall in 2009. House prices also fell in many other countries in 2008.
Many of thc ultimate investors in the big Wall Street institutions did not appreciate the riskiness ol their investments and their insurance obligations, and were shocked to see their complex securitised packages containing apparently safe-as-houses' mortgages become non-performing. As these failing mortgages began to unwind, they created losses on the asset side ol the balance sheets of financial institutions, as well as mounting liabilities due to their insurance obligations against defaults. These financial institutions were lelt with too much debt and too little capital to provide the rest of thc economy with much-needed credit. They tried to rescue their balance sheets by selling oil assets, but this forced down stock prices and intensified the problem. The widespread mortgage failures and thc subsequent crisis in interbank lending eventually led to the collapse ot some apparently rock-solid financial institutions.
The credit losses on mortgages are expected to continue into 2009. The IMF estimated in its 2008 World Economic Outlook that global losses on US mortgage failure will be about US$1.4 trillion, much more than Australia's entire C.DP: By October 2008, financial institutions had already written down US$0.8 trillion in their net worth. The banks lhat suffered the most were the investment banks, like Lehman and Merrill Lynch, whose liabilities included a significant quantity ol debt. AIG suffered because it was the major provider to banks of insurance against the default of their risky assets.
ТНт MACROFCONOMICS pi IINANC1AI MARKfcl CRISB
chapter 22
Financial markets depend critically on trusting that the banks they deal with and thc paper assets that they hold will not go into default. When lhat confidence disappears, linanciai institutions retreat Irom lending, trying to conserve precious capital. Credit markets can go into crisis when the only lending that financial institutions are prepared to do is ai very short horizons and at much higher interest rates. In 2008, interbank market interest rates increased significantly. With new loans becoming either impossible to get or very expensive, a system-wide credit crunch set in that hurt many households and lirms. The problem is that credit markets are at the frontline ol the day-to-day functioning of the economy with many businesses depending on short-term borrowing to finance their regular expenditures.
FOCUS
'BOX
The TED spread is a good indicator ol the costliness ol short-term borrowing. This spread is thc difference between interbank rates that reflects the degree ol perceived market riskiness and the equivalent sate Treasury hil rates. When thc crisis hit, the TED spread jumped dramatically—fifteen- fold! More details on the TED spread are provided in the focus box An indicator of the credit crunch— the TED spread'.
AN INDICATOR OF THE CREDIT CRUNCH—THE TED SPREAD
A useful indicator of the credit crunch is known as the 'TED spread'. This measure subtracts the rate of interest on a three-month US treasury bill (the T in TED) from the three-month eurodollar rate ('ED') measured by the three-month US$ London Inter-Bank Offer Rate (LIBOR). The TED spread indicates the extent of credit risk, since the LIBOR includes a premium for the default risk from lending between commercial banks, while the Treasury bill is considered to be risk free.The TED spread rises if the risk of bank defaults increases. Figure I presents the TED spread from 2003 to September 2008. The spread rose dramatically from August 2007. when the sub-prime mortgage problem first began to unfold. By October 2008 the spread had reached 450 basis points, fifteen times greater than it had been prior to 2007. Central banks and Treasuries nervously watched indicators like TED, but it was only in September 2008 that they began to
Figure I
The TED spread.
2003-08
The TED spread measures the difference between the three-month LIBOR interest rate and the three-month US Treasury bill rate. An increase signifies greater bank credit risk. From mid-2007 the spread jumped from an average of about 30 basis points to reach more 4S0 in October 2008.
SOURCE: FRED (Federal Reserve Bank of St Louis),
о a
1 J3
£
take serious action.The huge TED spread in late 2008 suggests that the key problem in financial markets was the perceived risk of insolvency, rather than insufficient money. This implies that the conventional monetary policy response of lowering the overnight interest rate may not have a big effect in the short term when the crisis is acute, and why we saw the Treasury (in the United States and other countries) intervening to try to re-capitalise the banking system and to provide government guarantees for deposits and for interbank claims.
Thc IMF forecast in 2008 that ihe crisis would lead to a significant write-down ol global bank capital and would slow growth in industrialised economies by at least 1.5 percentage points.
Regulation failure
Why didn't financial regulators respond to the growing sub-prime mortgage crisis: Unfortunately, mortgage finance regulation by the US authorities had been missing the mark (or years. The financial regulatory system in the United States was essentially designed in the 1930s alter the Creai Depression. Ii targeted regular commercial banks, which now play only a small part in housing finance, and ignored the large investment banks that held a significant amount of mortgage-backed securitised bonds, and which wrote large volumes ol insurance contracts against default of these bonds. By 2008, many people were profitably investing their funds in investment banks, which were lightly regulated because their deposits had no government guarantees. They didn't appreciate the risks.
Financial regulation needs to be a lot smarter in the future, and in 2009 President Obama will have to sort out the regulatory problems in US housing finance. This is a challenge lacing many countries whose financial systems have evolved dramatically in recent years. The regulatory framework must keep pace with financial system development. An important changing feature of the financial environment in thc 2008 crisis was the number of failures, mergers and takeovers ol financial
institutions. This meant that the industry became more concentrated, and so smarter regulation is even more essential.
Government policy responses
Though it is vitally important to intervene to prevent a house-ol-cards collapse ol the financial system, economists worry about the longer-term problem of moral hazard—it financial institutions come to expect that the government will always hail them out in times ol trouble, thev will have an incentive to take on more risk than they otherwise would. We discuss this issue further in the next chapter's locus box The Japanese banking problem'. Manv people are also uncomfortable with the idea of the taxpayer subsidising the losses incurred by linancial executives and shareholders who were paid handsomely when their industry was doing well Indeed when Lehman Brothers, a renowned Wall Street investment bank, indicated that it was insolvent in mid-September 2008, US policy-makers decided to let it tail. Many think that this failure was a major catalyst lor thc subsequent turmoil in financial markets.
Seeing the turn ol events, then Treasury Secretary Paulson finally decided to act rather than talk. The risk of another Great Depression was felt to be too great, and so with the support of Governor Bernanke of thc ГГО who had been actively responding to the crisis from thc outset, starting with Bear Stearns in March 20081 Paulson negotiated to commit government funds to protect some ol the financial institutions from failure. The US Senate and the US Congress agreed in October 2008 to provide US$700 billion in taxpayer lunds to buy out non-performing mortgages from troubled linancial institutions the so-called Troubled Assets Relief Program . or TARP lor short l. with thc intention of selling them back to the institutions in the luture when the market settled down. Some economists estimate that the rcscuc may end up costing thc US taxpayer more than US$ I trillion! 1 lowever, there is also a chance that thc crisis will eventually resolve with no cost to the taxpayer, and possibly even a large profit.
1 hough the financial crisis originated in US housing finance, il has spread like a contagious disease to the rest of the world. Policy-makers in Lurope and Asia have been forced to take the same sort of action as their counterparts in the United States. Why has the crisis spread? The reason is the intcrconnectedncss ol financial markets across thc world, l inancial markets are highly globalised and they provide the extremely valuable function ol pooling and spreading risk across the world. This allows households and firms to gel greater access to linance. enhancing iheir investment in new physical capital, new products and new technologies. Thai is one of the benefits of a well-functioning financial system. But it has its downside too. With risks spread across the globe, trom Iceland to Japan. Irom one continent to another, a major insolvency in Wall Street will be Iclt everywhere.
On 8 October 2008. six major central banks (in the United States, Europe, the United Kingdom, Canada, Sweden and Switzerland) recognised thc seriousness of the crisis by coordinating a cut in their policy interest rates by 0.5 percentage points. The Bank of lapan indicated its support but it was helpless to cut its rate because thc rate was already 0.5 percent. The Bank of China cut its rate that day by 0.27 percentage points. Policy interest rates are sure to be cut more in 200У, but the rate in many countries is last approaching the limit, zero. At this limit, the economy enters whai is known as the liquidity trap, an issue we will discuss further in thc next chapter.
4 Here's a joke that Icelanders won't appreciate: What's the capital of Iceland? $9.99!
THE MACROECONOMICS OF FINANCIAL MARKET CRISIS
chapter 22
©
In Australia the Reserve Bank (RBA) was so concerned about the damage thai might be done to the real economy that il dropped the cash raie by a full 1 per ccnt to 6 per cent on 7 October 2008. even though it expected thc inllation rate to rise lo 5 per cent in early 2009, well above its inllation larget range, The RBA realised lhat such a large unexpected cut would lead to a major lall in thc exchange rate (and it did), but it felt lhat its strong action would help instil conlidcncc over time in financial markets. More large cuts in the cash rate are cxpectcd. Thc government declared lhat all banks in Australia were in good linancial shape. However, because banks in other countries were receiving government-backed deposit guarantees, Australian banks needed the same competitive edge. The government has thus guaranteed all deposits at Australian banks (supervised by the bank regulator, AI'RA) until 2011. Including a sunset clause on government guarantees is a smart move, since il puts a ceiling on the moral hazard problem.
Recognising that monetary policy has its limits the Rudd government decided in October 2008 to abandon its previous commitment to liscal surpluses, explaining that it was necessary in good times to build up these surpluses so that reserves would be there to be deployed in bad times. As a consequence, it introduced a $10.4 billion liscal expenditure plan (entitled the Economic Security Strategy ■ to try to increase aggregate demand in the economy. It is lar easier (or Australia to run tiscal delicits than it is tor many other countries. This is because thc Australian government is one of the lew that has negative net debt. Nevertheless, there is little doubt that tiscal expansion needs to provide a central role everywhere and that more will be needed than is currently proposed.
At the time of writing October 2008), it is unclear whether thc proposed massive interventions by policy-makers around the world will work—whether they will be enough to prevent the crisis Irom creating a recession, or even a depression. Many politicians and government officials are not sleeping too well right now-
One key lesson to learn from financial crises is that policy-makers need to keep an ever-watchlul eye on developments in the financial markets. These markets function well when the level ol trust is high. That trust cannot be legislated lor. When it disappears, a linanciai crisis will occur.
When linanciai crises are long and deep, thc macroeconomic pathology becomes much more serious. In the next chapter we look at the Great Depression in the 1930s and at the long lapanese slump that began in the early 1990s.
SUMMARY
• Financial crises have occurred on average every ten years over the last two centuries.
• Financial institutions provide a valuable intermediation function between savers and borrowers. Over the last quarter ot a century, linanciai innovation has been remarkable helping to spread risk and enhance economic activity.
• Rut sometimes the financial system becomes stressed. When asset prices start tailing, bank balance sheets suffer. In a linanciai crisis, banks' net worth lalls, and they risk insolvency. They may try to sell assets and so asset prices lall even more. Further, banks restrict their lending to each other because of the tear ot default. Interbank interest rates rise significantly, rellecting the higher dclault risk.
• Lower asset prices and higher interest rates on borrowing affect the real economy by reducing real investment and consumption. Output tails The central bank should lower its policy interest rate by the amount that the credit risk premium rises, but it cannot lower it below zero. Pessimistic future expectations may cause output to fall even further.
• The global linanciai crisis ol 2008 began with problems in housing linance in the Llnited States. When house prices started to fall in 2007, many home-owners with sub-prime mortgages foreclosed on their loans. These delaults severely alfected the balance sheets ol many investment banks on Wall Street a number ot which tailed merged or were taken over.
• The effects spread beyond the Llnited States, with share prices falling everywhere, and credit becoming very hard to obtain.
• Policy-makers responded by lowering interest rates, providing guarantees lor deposits injecting capital into struggling banks, and rethinking the regulation of financial institutions.
• As monetary policy reaches its zero interest rate limit liscal policy expansion begins to play a central role in trying to avoid the risk ot a fully fledged depression.
THL MACROECONOMICS Ot FINANCIAL MARKET CRISES
chapter 22
KEYTERMS
financial intermediation, 515 • credit default risk premium. 517
securitised bonds, 515 • sub-prime mortgages, 520
interbank interest rate, 516 • TED spread, 521
credit crunch, 516 • moral hazard, 523
deposit guarantees, 516
QUESTIONS AND PROBLEMS
Quick check
1. Using the information in this chaptcr, label each of lhe following statements 'true', 'false' or 'uncertain'. Explain briefly.
a. Financial institutions are intermediaries that match savers and borrowers.
b. Financial institutions provide a valuable service in spreading risk.
c. Debt ot any form on a balance sheet is always bad.
d. The central bank can always lower the interest rate it controls by enough to counter any credit crisis.
c. Policy-makers should think about the problem ol moral hazard when they protect banks from failure.
2. Bank balance sheets
Using the simplified balance sheet structure given in Table 22.1, consider the bank management's response to:
a. a bank run on its deposits
b. a decrease in the market value ot its loans and bonds
c. an increase in its debt obligations
d. a collapse in its own share price
3. Policy responses in a financial crisis
Using the IS-LM model with expectations of future variables developed in Chapter 17, explain how policy-makers might respond to the following shocks:
a. A collapse in consumer confidence.
b. A lall in stock-market prices expected to correct itself within weeks.
c. A permanent tall in aggregate demand, and thus the natural level ot output.
d. A permanent lall in thc natural level ol output, and a temporary but persistent increase in the market interest rate (due to a rise in the default risk premium).
Dig deeper
4. Write a brie! but critical summary and comparison ol the key policy responses to the 2008 financial crisis hy policy-makers in the United States, Europe, Japan and Australia. Did any of these countries lace the zero nominal interest rate constraint?
5. 1 low deep and long was the effect of the 21)08 crisis?
Collect quarterly data on unemployment and output growth from 2007 onwards for Australia, the United States, the United Kingdom and Japan. You can gel these from a number of places, such as the OECD, the ITS at the IMP, the RBA, The Economist website,
Economagic website,
a. Put the output growth scries tor all the countries on a graph. Do the same lor unemployment.
b. Which country appeared lo suffer the most? The least? Give reasons.
c. Which country appeared lo suiter for the longest lime? The shortest lime? Give reasons.
Explore further
6. TED spreads
A key indicator of the cost of borrowing by banks, firms and households in the market is the spread between a risky market interest rate and an equivalent safe government bond.
Construct daily TED spreads for the month of October 2008 AND the most recent month available for Australia, the United States, the United Kingdom and Japan. For the risky market rates, you can use the LIBOR rates for three-month maturities in the -various currencies, available on
a. Which of the country's spreads is the highest?
b. Has the ranking changed?
c. looking back at your answer to thc previous question, was the TED spread a reasonable prcdictor ol ihe relative pertormance of the countries in terms ol output growth and unemployment?
We invite you to visit the Blanchard-Sheen page on the Pearson Australia website at
www.pearson.com.au/highered/blanchardsheen3e
for many World Wide Web exercises relating to issues similar to those in this chapter.
FURTHER READINGS
Robert Shiller explains the origins and possible cures for the sub-prime mortgage crisis in The Subprime Solution: How Today's Global Financial Crisis Happened, and What lo Do About It (Princeton, NJ: Princeton University Press, 2008).
An excellent survey of the current 2007-08 linancial crisis can be found in When fortune Irowned. A special report on the world economy', The Economist, 1 I October 2008.
Thc IMF's World Economic Outlook in 2008 provides a timely analysis ot the 2008 crisis, particularly in Chapter 4,
For a readable discussion of the issues, have a look ai some ol the papers in The Economist's Voice, including the September 2008 issue on 'Financial regulation, financial crisis, and bailouts',
Thc RBA's annual confcrcnce in 2008 was on the financial crisis of 2007-08. You can find thc papers on the conference section of their website,
For an analysis ol the policy challenges lacing governments in thc global financial crisis, read Olivier Blanchard, 'Cracks in ihe system: Repairing ihe damaged global economy', Finance and Develop¬ment, December 2008,
CHAPTER
Depressions and Slumps
A
major theme of this book so far has been that, while economies go through fluctuations in the short run, they tend to return to normal in the medium run. An adverse shock may lead to a recession, but fairly quickly the economy turns around and output returns to its natural level.
Most of the time, this is what happens. But once in a while things go seriously wrong. Output remains far below its natural level for many years. Unemployment remains stubbornly high. Simply put, the economy appears to be stuck, unable to return to normal.The most infamous case is surely that of the Great Depression, which affected most of the world from the late 1920s to the start of World War II. (While there is no agreed-upon definition, economists use the term depression to describe a deep and long-lasting recession.) A more recent case is the Japanese slump that started in the early 1990s. from which Japan is slowly emerging. (Again, while there is no agreed-upon definition, most economists use slump to denote a long period of low or no growth, longer than a typical recession but less deep than a depression.) What goes wrong during such episodes? Are the shocks particularly bad? Do the usual adjustment mechanisms break down? Or are macroeconomic policies particularly misguided? These are the questions we take up in this chapter.
• Section 23.1 looks at two of the mechanisms that have played a central role in the Great Depression and in Japan: the adverse effects of deflation, and the liquidity trap.
• Section 23.2 gives an account of the Great Depression.
• Section 23.3 does the same for the Japanese slump.
23.1 DISINFLATION, DEFLATION AND THE LIQUIDITY TRAP
Low output leads to о decrease in the price level. The decrease in the price level leads to on increase in the real money stock. The LM curve shifts down, and it continues to shift down until output has returned to its natural level.
rate and higher output. Alter some time, the economy will be, for example, at point B. with output equal to У'.
So long as output remains below its natural level, thc pricc level will continue to decrease, and the LM curve will continue to shift down. The economy will move down the IS curve until it reaches point С and output has returned to Y„. In short, output below the natural level ol output will lead to a decrease in the price level, which will continue until the economy has returned to normal. This argument in Chaptcr 7 was based on the strong simplifying assumption that the nominal money >ilock is held constant. This implies that in the medium run the price level is also constant. And it also implies that, if output is below its natural level, the adjustment ol output back to its natural level is achieved through a dccreasc in the price level—something wc rarely observe in practice. Chapters 8 and 9 explored a more realistic story, where we allowed lor positive inflation—and so allowed for positive nominal money growth in the medium mn. This model gave a richer description of the adjustment of output and inflation to shocks. But, lor our present purposes, it delivered thc same basic implication as the simpler version ol the model presented in Chapter 7: the economy will tend to return lo the natural level ol output over time.
The argument now goes as follows:
Suppose that as in Figure 2.3.1, output is below its natural level—equivalently, the unemployment rate is higher than its natural rate. Then, Irom the Phillips curve relation inllation will decrease over time.
Suppose that nominal money growth and inflation were initially equal to each other, so that real money growth the difference between nominal money growth and inllation) was initially equal lo zero.
For given Ad: ►
Pi т-ь M/Pt =» LM
Г h iff С HriVA/n V
From equation (8.10): If ► the unemployment rate exceeds its natural rate, inflation decreases.
LM
Figure 23.1 The return of output to its natural level
II inflation decreases, and so becomes lower than the rale ol nominal money growth, then real money growth will now turn positive. Equivalently, the real money stock will increase.
• This increase in the real money slock will shilt the LM curve down, leading to an increase in output. The LM curve will continue lo shilt down until, eventually, output is hack to its natural level.
So, ihe adjustment looks the same as in Figure 23.1: lower output will lead to an increase in thc real money stock, until output is back to its natural level.
It would therefore appear that, even if the ccntral bank holds money stock growth constant, economies have a strong built-in stabilising mechanism ю lili them out ol recessions:
• Output below its natural level leads lo lower inflation.
• Lower inflation leads in turn to higher real money growth.
• Higher real money growth leads to an increase in output over lime.
4 Let r be the real interest rate, i be the nominal interest rate and тс be expected inflation.Then, from equation (14.4): r=i-7rc.
* r = i - if. So. for a given I. u*Wr=*y.l.The IS curve shifts to the left.
In Chapter 9 you saw that, in the medium run. inflation is equal to nominal money growth minus normal output * g'Owth.The example assumes for simplicity that normal output growth is zero, so inflation and nominal money growth are equal.
Thc study ol depressions and slumps tells us. however, thai this built-in mechanism is not foolproof, and that things may go wrong in several ways. We now look at some ol these ways.
The nominal interest rate, the real interest rate and expected inflation
Looking al ihe adjustment ol output in Figure 23.1, we ignored the distinction between the nominal interest rate and the real interest rate. We need now lo reintroduce ihis distinction. Recall from C hapter 14 that:
• Whai matters lor spending decisions and thus what enters the IS relation, is thc real interest rate— the interest rate in terms ol goods.
• What mailers lor the demand lor money, and thus what enters ihe LM relation, is ihe nominal interest rate—ihe interest rate in terms ol dollars.
Recall also thc relation between thc two interest rales-, thc real interest rate is equal to the nominal interest rate minus expected inflation.
What this distinction between thc two interest rates implies is shown in Figure 23.2. Suppose that ihe economy is initially at A: output is initially below its natural level. Because output is below its natural level, inflation decreases.
• The decrease in inllation leads to an increase in the real money slock and a shift in the LVI curve down, trom LM to LM'. Thc shitt ot thc LM curvc—due to thc increase in MIP—is thc shilt we saw in Figure 23.1. This shili ol the L.V1 curve tends to increase output. II this were the only shift, the economy would go from A to B.
• But there is now a second effect ai work. Suppose the decrease in inllation leads to a dccrcasc in expected inflation. I hen. tor a given nominal interest rale, the decrease in expected inllation increases thc real interest rate. The higher real interest rate leads in lurn to lower spending and lower output. So. at a given nominal interest rate, ihe level ot output implied by equilibrium in the goods market is lower. The /5 curve shifts to ihe left, from IS to IS'. Thc shift in the IS curvc—due to the decrease in 7Г'—tends to decrease output, li this were the only shilt, thc economy would go from A to B'.
Does output go up or down as a result of these two shitis? Wc cannot tell. The combined cffcct ot the two shifts is to move the economy trom A to B", with output У". Whether Y" is greater or smaller than V depends on which shili dominates, and is in general ambiguous.
As we have drawn the :igure V" is smaller than V. In this ease raiher lhan returning to its natural level output declines further away Irom it: things get worse rather than belter.
A numerical example will help you to keep straight thc two effects ol inllation on output:
• Suppose lhat nominal money growth, inflation and cxpectcd inllation arc all equal initially lo 5 per ccnt.
Suppose lhat ihe nominal interest rale is equal to 7 per cent, so the real interest rale is equal
to 7% - 5% = 2%.
4 Assume that thc rate of nominal money growth, g„. and the rate of inflation. - are initially equal: gm = jr. Then, for given g„„ -I —>
g„-->o=>
MlPT => LM shifts down.
rt.
• Suppose thai, because output is lower than its natural level, inllation decreases from 5 per ccnt lo 3 per cent alter a year.
Figure 23.2 The effects of ■ LM
lower inflation ■
on output о
ы
2 *J
И
2! 0
б .£ i \ Атгв < 0 4 \ Л ■ ■ ■
к
/■ ■
■
■
1 1
■ /
1 /
/ 1 f | / 1 S(MIP) > 0 / Ш
"я \ \ в/
11 ■
1 О
Z \в 1
1 ■
ч,. ■ ■
sjs- : ^^^ а IS
У" У
Output, У
When inflation decreases in response to low output there are two effects:
(1) The real money stock increases, leading the LM curve to shift down.
(2) Expected inflation decreases, leading to a shift of the IS curve to the left The result may be a further decrease in output
The liquidity trap
One reaction to the scenario just described is to conclude that, while we should worry about it, it can easily be avoided by the appropriate use ol macroeconomic policy—in particular, monetary policy. Thc scenario was derived under the assumption that monetary policy (in our case, the rate ol growth of nominal money) remained unchanged. But if the ccntral bank is worried about a decrease in output, it would seem that all it needs to do is to embark on an expansionary monetary policy. In terms of Figure 23.2, all the central bank needs to do is to increase the stock of nominal money to shilt thc LM curve down further, and make sure that the shilt in thc LM curve will be enough to increase output.
That monetary policy can and should be used in this contcxt is clearly thc right prescription. But there is a limit to what the central bank can do. It cannot decrease the nominal interest rate below zero, an issue wc raised in the last chaptcr. Il cxpcctcd inflation is low or even negative (if people expect a deflation), the implied real interest rate may still not be low enough to get the economy out of a recession. This issue is at the centre ol discussions about Japan today. Let's now look at it more closely.
Go back lirst to our characterisation ol the demand and the supply of money in Chapter 4. There, we drew the demand for money, for a given level of income, as a decreasing function ol thc nominal interest rate. The lower the nominal interest rate, the larger the demand for money—equivalently, the smaller the demand tor bonds. What we didn't ask in Chapter 4 is what happens when the interest rate goes down to zero. 1 he answer: Once people hold enough money for transaction purposes, they arc then indifferent to holding the rest ol their linancial wealth in the form ol money or bonds. The reason they are indifferent is that money and bonds pay the same nominal interest rate—namely, zero. Thus, the demand for money is as shown in Figure 23.3:
• As the nominal interest rate decreases, people want to hold more money and thus fewer bonds). The demand for money increases.
4 Look at Figure 4.1 in Chapter 4. We avoided the issue by not drawing the demand for money for interest rates close to zero.
• As the nominal interest rate becomes equal to zero people want to hold an amount of money at least equal to the distance OB this is what they need tor transaction purposes But they arc willing to
Figure 23.3 Money demand, money supply and the liquidity trap
As the nominal interest rate decreases to zero, once people have enough money for transaction purposes they are indifferent between holding money and holding bonds. The demand for money becomes horizontal. This implies that, when the nominal interest rate is equal to zero, further increases in the money supply have no effect on the nominal interest rate.
РАГноюакз > ■ chapter 23
hold even more money (and therefore hold lewer bonds) because they are indifferent between money and bonds. Therefore, the demand lor money becomes horizontal beyond point B. Now consider the ellects ol an increase in the money supply.
Consider the case where the money supply is M\ so the nominal interest rate consistent with financial market equilibrium is positive and equal to /. 'This is thc case we considered in Chapter 4.) Starting Irom that equilibrium in Figure 23.3, an increase in the money supply—a shilt ol the AT*' line to the right—leads to a decrease in the nominal interest rate.
Now consider the case where the money supply is AT"', so the equilibrium is at point В■ or consider the case where the money supply is AT", so the equilibrium is given at point C. In either case, the initial nominal interest rate is zero. And in either case an increase in the money supply has no effect on the nominal interest rate. Think ol it this way: Suppose the central bank increases the money supply. It docs so through an open-market operation in which it buys bonds and pays for them by creating money. As the nominal interest rate is zero, people are indifferent to how much money or how many bonds they hold so they are willing to hold fewer bonds and more money at the same nominal interest rate—namely, zero. The money supply increases but with no effect on thc nominal interest rate. In short, once the nominal interest rate is equal to zero expansionary monetary policy becomes bonds in exchange for powerless. Or to use the words ol Keynes, who was the lirst to point to ihe problem, the increase in money money lalls into a liquidity trap People are willing to hold more money more liquidity at the same nominal interest rate.
From Chapter 4:The ► central bank changes the money stock through open-market operations, in which it buys or sells
Having looked at equilibrium in the financial markets, lets now turn to the IS-LM and see how it must be modified to take into account the liquidity trap.
с о Z
Md" (forY" < Y)
Md' (for У' < У)
MIP
(Real) money, MIP
(a)
Md (for given income, У) M*
Income, У
(Ь)
The derivation ol the LAI curve is shown in Figure 23.4, panels (a) and (b). Recall that the LAI curve gives, lor a given real money stock, the relation between the nominal interest rate and the level of income implied bv equilibrium in financial markets. To derive the LM curve, panel a) looks at
Figure 23.4 The derivation of the LM curve in the presence of a liquidity trap
«
с
Ё о Z
LM curve
For low levels of output, the Lli curve is a fat segment, with a nominal interest rate equal to zero. For higher levels of output, it is upward sloping: an increase in income leads to an increase in the nominal interest rate.
equilibrium in ihe financial markets for a given value ol ihe real money stock, and three money demand curves, each corresponding ю a different level of income-.
• Md gives the demand lor money lor a given level ol income V. Thc equilibrium is given by point A, with nominal interest rate equal to i. This combination of income Vand nominal interest rate i gives us a lirst point on the LM curve, point A in panel (b).
• M'1' gives the demand lor money lor a lower level ol income, У < У. Lower income means fewer transactions, and so a lower demand lor money at any interest rate. In this case, the equilibrium is given by point A', with nominal interest rate equal to i". This combination of income, У, and nominal interest rate, i', gives us a second point on thc LM curve, point A in panel (b).
• M1'" gives ihe demand for money for a siill lower level ol income, Y" < V. In ibis case, the equilibrium is given by point A" in panel (a), with nominal interest rate exactly equal to zero. Point A" in panel (b) corresponds lo A" in panel (a).
• Whai happens it income decreases below V", shifting the demand tor money further to thc left in panel (a)? The intersection between the money supply curvc and thc money demand curvc takes place on the horizontal portion of the money demand curve. The equilibrium remains at A", and the nominal interest rate remains equal to zero.
Let's summarise. In thc presence ol a liquidity trap, the LM curve looks as drawn in l igure 23.4, panel (b). Г:ог values ot income greater than Y" it is upward sloping—iust as il was in Chapter 5 when we first characterised the /./VI curve.
4 So far, the derivation of the LM curve is exacdy the same as in Chapter 5. It is only when income is below Y" that things become different.
Ф i
l or values of income less than У it is flat at i 0: thc nominal interest rate cannot go below zero. Having derived the LM curve in the presence of a liquidity trap, we can look at the properties of ihe IS-LM model modified in this way. Suppose that the economy is initially at point A in Figure 23.5. Equilibrium is at the intersection ol the IS curvc and the LM curve, with output, Y, and nominal interest raie. /'. And suppose that this level ot output is tar below the natural level ot output, Y„. The question is: Can monetary policy help the economy return lo Y„?
IS ■
а ■
а а Figure 23.5 The IS-LM model and the liquidity trap
Ф a 3 k.
w В 0) к i
A £ в в в в в
в ■
1 ■
в LM
LM'
LM"
rt с
£ 0 z
i \ A
В в в 1 1
в
■
в
в ■
:
: ■
■ / /
Y Y' Output, Y
In the presence of a liquidity trap, there is a limit to how much monetary policy can increase output Monetary policy may not be able to increase output back to its natural level.
Suppose that the central bank increases the money supply, shifting the LM curve from LM to LM'. The equilibrium moves from point A down to point B. lhe nominal interest rate decreases Irom i to zero, and output increases Irom V to Y'. Thus, to this extent, expansionary policy can indeed increase output.
What happens, however, il, starting from point B, the central bank increases the money supply further, shilling the LM curve Irom LM' to. say, LM"? The intersection ol IS and LM" remains at point /i and output remains equal to Y'. Expansionary monetary policy no longer has an effect on output; it cannot therefore help output return to Y„.
In words: When the nominal interest rale is equal to zero, the economy falls in a 'liquidity trap . The central hank can increase liquidity—that is. increase the money supply. But this liquidity' lalls into a trap. The additional money is willingly held by financial investors at an unchanged interest rate— namely, zero. II at this zero nominal interest rate the demand for goods is still too low, there is nothing further that monetary policy can do to return output to its natural level.
Putting things together:The liquidity trap and deflation
Just as you may have been sceptical when we were discussing the adverse effects ot lower inllation earlier, you may remain sceptical that the liquidity trap is a serious issue. Alter all, a zero nominal interest rate is a very low interest rate. Shouldn't a zero nominal interest rate be enough to strongly stimulate spending and avoid a recession?
The answer is no'. And to explain it. we must again draw the distinction between the real interest rale and the nominal interest rate. What matters for spending is the real interest rate. What the real interest rate corresponding to a zero nominal interest rate is depends on thc rate of expected inflation:
• Suppose the rate ot inflation, actual or expected, is high—say, equal to 10 per cent. Then, a zero nominal interest rate corresponds to a real interest rale ol —10 per cent. At such a negative real interest rate, consumption and investment spending are likely to be very high high enough to make sure lhat demand is sufficient to return output to the natural level of output. So. at high inflation the liquidity trap is unlikely lo be a serious problem.
• Suppose lhat the rate ol inflation is negative—thc economy is experiencing deflation. Say, the rate ot inflation is -5 per cent equivalently, the rate ol deflation is 5 per cent). Then, even if the nominal interest rate is equal lo zero, the real interest rate is equal to 5 per cent. This real interest rate may still be too high to stimulate spending enough, and, in this case, there is nothing monetary policy can do to increase output.
You can now see how the two mechanisms—the effects ol expected inflation on the real interest rale, and the liquidity trap)—described in this section can come together to turn recessions into slumps or depressions.
Suppose that the economy has been in a recession for some time, so inflation has steadily decreased and turned into deflation. Suppose that monetary policy has decreased the nominal interest rate down to zero. Even at this zero nominal interest rale, expected deflation implies that the real interest rate is still positive.
Suppose that, as a result, the economy is at a point such as /1 in Figure 23.6. at the intersection ol the IS and the LM curves. The nominal interest rate is equal to zero, and output, V, is below ihe natural level of output. Y„. There is clearly nothing monetary policy can do in this case to increase output. And things are likely to get worse over time.
As output is below its natural level, thc rate ol deflation, actual and expected, is likely to increase. (Inflation is likely to become more negative. At a given nominal interest rate, higher expected deflation leads lo an increase in the real interest rate, the IS curve shifts to the left in Figure 23.6, from IS to, say, IS', leading to a further decrease in output, from V down to Y'.
Even if the central bank ► chooses to fix the interest rate (rather than the money stock) based on an inflation target, it can fall into a liquidity trap.The trap occurs when the nominal interest rate reaches zero.
г = ,- TT' = 0%-I0%> = -10%
Revisit the discussion of investment decisions in
Chapter 16. Why is ^ investment likely to be very high if firms can borrow at a real interest rate of-10 per cent? (Hint What do firms compare the real interest rate win?)
r = i - rc = 0% - (-5%) = 5%
This leads to lurther deflation, which leads to a further increase in the real interest rate, a further decrease in output, which leads . . . and so on.
Figure 23.6 The liquidity trap arid deflation
Suppose that the economy is in a liquidity trap and there is deflation. Output below its natural level leads to more deflation over time, which leads to a further increase in the real interest rate, and to a further shift of the IS curve to the left This shifts leads to a further decrease in output which leads to more deflation ... and so on.
Warning: The quality of unemployment data is
^ much lower before World War II than after it Cross-country comparisons are particularly dangerous.
4 For a look at other countries, read Peter Temin s Lessons from the Great Depression (Cambridge. MA: MIT Press, 1989).
The economy gets in a vicious cycle. Low output leads to more deflation. More deflation leads to a higher real interest rate and even lower output, and there is nothing monetary policy can do about it. Thc scenario may sound exotic, which it may be. but, as we will see when looking first at the Great Depression and then at thc lapanesc slump, it is lar from irrelevant.
23.2 THE GREAT DEPRESSION
Figure 23.7 The US unemployment rate. 1920-50
30
25-
20-
15-
5-
v
0
5
*j
z
V
E
1
a.
E
V
с D
10'
"I—I—I—i—i—i—i—i—i—(—1—I—i—I—I i—i—i—i—i—r—l—I
1920 1922'1924'1926 1928 1930 1932 1934 1936 1938 1940 1942 1944 1946 1948 1950
From Section 9.1: Okun's law relates the change in the unemployment rate to the deviation of output growth from nornal output growth. In the United States today, outout growth of I per cent above normal for a year leads to a decrease in the unemployment rate of about 0.4 per cent. If normal output growth is 2 per cent, show, using Table 23.1. how this quantitative relation fits the relation between output growth ^ and unemployment from 1933 to 1941.
The Great Depression was characterised by a sharp increase in unemployment, followed by a slow decline.
Table 23.1 Unemployment, output growth, prices and money. United States, 1929—42
Unemployment Output growth Nominal
Year rate (%) rate (%) Price leveL money stock
1929 3.2 -9.8 100.0 26.4
1930 8.7 -7.6 97.4 25.4
1931 15.9 -14.7 88.8 23.6
1932 23.6 -1.8 79.7 19.4
1933 24.9 9.1 75.6 21.5
1934 21.7 9.9 78.1 25.5
1935 20.1 13.9 80.1 29.2
1936 16.9 5.3 80.9 30.3
1937 14.3 -5.0 83.8 30.0
1938 19.0 8.6 82.2 30.0
1939 17.2 8.5 81.0 33.6
1940 14.6 16.1 81.8 39.6
1941 9.9 12.9 85.9 46.5
1942 4.7 13.2 95.1 55.3
1 1
SOURCES. Unemployment rate: Series D85-86; output growdv GNP growth (in 1958 prices). Series F31: price level: CPI (1929 = 100). Series E135; money stock: Ml (in bill ons of dollars). Series X414. Historical Sinusites of the United Stales. US Department of Commerce.
a high 9.9 per cent. (There is no contradiction here, just an application ot Okun's law: a long period of high growth was needed to decrease thc unemployment rate. Let's look at these two aspects in turn.
The initial fall in spending
Popular accounts often say that the Great Depression was caused by the stock-market crash of 1929. Not so. A recession had actually started before the crash, and other factors played a central role later in the depression.
Nevertheless, the crash was important. The stock market had boomed from 1921 to 1929. Slock prices had increased much faster than the dividends paid by firms—and, as a result, the dividend-price ratio had decreased from 6.5 per cent in 1921 to 3.5 per ccni in 1929. On 28 October 1929. the stock- market price index dropped Irom 298 to 26(1. The next day, it dropped to 230. This was a lall of 23 per cent in two days, and a drop ot 40 per cent from the peak of early September. By November the index was down to 198. A bricl recovery in early 1930 was followed by further declines in slock prices as the depth ol the depression became increasingly clear to the market participants. In June 1932 the index bottomed out, at 47. (The evolution ol the index from lanuary 1920 lo December 1950 is shown in Figure 2.3.8.1
Was the October 1929 crash caused by the sudden realisation that a depression was coming? The answer is 'no'. There is no evidence of major news in October. The source ot the crash was almost surely thc end ol a speculative bubble. Stockholders who had purchased stocks at high prices in the anticipation of further increases in prices became scared and attempted to sell their stocks. Thc result was a large drop in prices.
320-
280-
240-
200-
160-
120-
l—l—1—l—l—I—I—I—l—I—I—I—I—1—I—l—l—I—I—I—I—I—I—I—I—I—I—I—I—I—I
1920 1922 1924 1926 1928 1930 1932 1934 1936 1938 1940 1942 1944 1946 1948 1950
v с
Figure 23.8 The US S&P composite index. January 1920 to December 1950
From September 1929 to June 1932 the stock-market index decreased from 313 to 47. and slowly recovered thereafter.
The crash not only decreased consumers' wealth but also increased their uncertainty about thc future. Unsettled by the crash, consumers and firms decided to see how things evolved and to postpone purchases ol durable goods and investment goods. There was, lor example, a large decrease in car sales the type ol purchase that can easily be delerred—in the months just lollowing the crash. Industrial production which had declined by 1.8 percent from August to October 1929, declined by 9.8 percent from October ю December, and by another 24 percent from December 1929 to December 1930.
The contraction in nominal money
4 Note the parallel between the evolution of stock prices during the 1920s and during 2002-08. The US dividend-price rat о averaged about 1.6 per cent from 2002 to 2007. then jumped up to about 3 per cent in late 2008.
4 Revisit dividends and prices, and bubbles and crashes, in Section 15.3.
The impact ol the crash was compounded by a major policy mistake—namely, a large decrease in the nominal money stock. The lirst column ot Table 23.2 gives the evolution ol thc nominal money slock,
Table 23.2 Money, nominal and real. United States. 1929-33
Year Nominal money stock
(Ml) Monetary base
(H) Money multiplier (Ml/H) Real money stock (Ml/P)
1929 26.4 7.1 3.7 26.4
1930 25.4 6.9 3.7 26.0
1931 23.6 7.3 3.2 26.5
1932 20.6 7.8 2.6 25.8
1933 19.4 8.2 2.4 25.6
SOURCES: MI: Series X414: H: Series X422 plus Series X423: P: Series E135. Historical Statistics of thc United States. US Department of Commerce
measured by Ml. (Ml is the sum of currency, travellers' cheques and chcckablc deposits, i From 1929 to 1933, Ml decreased from US$26.4 billion to US$19.4 billion, a decrease of 27 per cent.
To understand why thc nominal money stock went down so much you must go back to what you learned in Chapter 4 about the relation between thc nominal money stock and the monetary base. In an economy in which some of the money held by people and lirms lakes the form of chcckablc deposits, the money stock (the sum of currency and checkable deposits) is larger than the monetary base, II (currency plus banks' reserves). The relation between the two is given by
MI H X Money multiplier
The money multiplier depends in turn both on how much resen'es banks keep in proportion to their deposits and on what proportion of money people keep in the form of currency as opposed to checkable deposits. Now note that trom 1929 to 1933 the monetary base, 11 (shown in the second column of Tabic 23.2). increased from US$7.1 billion to US$8.2 billion. This means that the decrease in Ml didnt comc Irom a decrease in the monetary base, but came instead from a decrease in the money multiplier, Ml/H (shown in thc third column of Table 23.2', which fell from 3.7 in 1929 to 2.4 in 1933. Why did the money multiplier decline so much? The answer: Because ol bank failures.
With the large decline in output, more and more borrowers found themselves unable to repay their loans to banks, causing more and more banks to become insolvent and close down. Bank lailurcs increased steadily from 1929 until 1933, when thc number of failures reached a peak ot 4,000, out ot about 20,000 banks in operation at the time.
Bank failures had a direct effect on the money supply. Checkable deposits at the failed banks became worthless. But the major cflect on the money supply was indirect: worried that their bank might also fail, many people took their money out ol the banks and shitted from checkable deposits to currency. The increase in thc ratio of currency to deposits led lo a decrease in ihe money multiplier, and so to a decrease in the money supply. Think of the mechanism this way: il people had liquidated all their deposits and asked banks tor currency in exchange, the multiplier would have decreased all thc way down to I. People would have held only central bank money,- M I would have been just equal to the monetary base H. Thc actual shift was less dramatic,- nevertheless, the multiplier dropped from 3.7 in 1929 to 2.4 in 1933, leading to a dccrcasc in the money supply despite an increase in thc monetary base.
What follows relies on the presentation of equilibrium in financial markets in terms of (overall) money supply and (overall) money demand presented in Chapter 4. ►
Checkable deposits in t the United States are equivalent to current account deposits with cheque book facilities in Australia.
The classic description t of what happened then is by Milton Friedman and Anna Schwartz. A Monetary History of the United States, /867-/960 (Princeton. NJ: Princeton University Press. 1963).
From Chapter 4: The multiplier is 1/(c + в (1 - с)) where с is the proportion of money ► people want to hold as currency, and H is the ratio of reserves to bank deposits. The higher с is. the lower the multiplier. And if с = 1 (if people want to hold only currency), the multiplier equals 1.
The implication for our purposes is simple: with a decrease in the nominal money stock Irom 1929 to 1933 roughly proportional to the decrease in the price level, the real money stock (shown in the fourth column of Tabic 23.21 remained roughly constant, eliminating one of thc mechanisms that could have led to a recovery In other words, the LM curve remained roughly unchanged—it didn't shift down as it would have done if the nominal money stock had remained constant, implying an increase in the real money stock.
This is why Milton Friedman and Anna Schwartz have argued that the Fed was responsible lor the depth of the depression. It wasn't directly responsible for the decrease in the nominal money supply, but it should have taken steps to oflset the decrease in the money multiplier by expanding thc monetary base much more than it did.
The adverse effects of deflation
With the lall in spending, and the decrease in the nominal money supply, the stage was set tor the mechanisms we studied in Section 23.1 to turn the decline in output into a fully Hedged depression.
As shown in the first column ol Table 23.3, the result ol the contraction in nominal money was to lead to only a limited decline in the nominal interest rate. The nominal interest rate measured by the interest rate on one-year corporate bonds reached 5.3 percent in 1929 (up Irom II percent in 19281, only to slowly decline over time, reaching 2.6 per cent in 1933.
Christina Romer. 'What ended the Great Depression?' .Journal of Economic History. < December 1992. pp. 757-84.
At the same time, as shown in the second column ol Table 23.3, the result ol low output was a strong deflation, with the rate of deflation reaching 9.2 per cent in 193 I and 10.8 per cent in 1932! It we make the assumption that expected deflation was equal to actual deflation in each year, we can constmct a series for the real interest rate. This is done in the last column ol Table 2.3.3, and gives a convincing explanation lor why output continued to decline until 1933. The real interest rate reached 12.3 per cent in 193 1. 14.8 per cent in 1932 and still a very high 7.8 per cent in 1933. It is no great surprise that, at those interest rates, both consumption and investment demand remained very low, and the depression got worse.
The recovery
The recovery started in 1933. Except lor another sharp decrease in the growth rate ol output in 1937 (see Table 2.3.1 I. growth was consistently high, running at an average annual rate ol 7.7 per cent Irom 19.3.3 to 1941. Macroeconomists and economic historians have studied the recovery much less than they have studied the initial decline. And many questions remain.
One of thc factors that contributed to the recovery is clear. Following the election ol Franklin Roosevelt in 1932, there was a change in monetary policy and a dramatic increase in nominal money growth. From 1933 to 1941, the nominal money stock increased by 140 per cent, the real money stock by 100 per cent. These increases were due to increases in the monetary base, not in thc money multiplier. Christina Romer, an economic historian Irom the Llniversity ot California at Berkeley has argued that, il monetary policy had been unchanged Irom 1933 on. output would have been 25 percent lower than it actually was in 1937. and 50 per cent lower than it was in 1942. These are very large numbers. Even if we believe that the numbers overestimate the ellect ol monetary policy, thc conclusion that monetary policy played an important role in the recovery is slill surely warranted.
Table 23.3 The nominal interest rate, inflation and the real interest rate, United States, 1929 33
Year One-year nominal interest rate (%). i Inflation rate (%). rr One-year real interest race (%). r
1929 5.3 -0.0 5.3
1930 4.4 -2.5 6.9
1931 3.1 -9.2 12.3
1932 4.0 -10.8 14.8
1933 2.6 -5.2 7.8
SOURCES: Interest rate: Series X487-491: inflation rate constructed from C'l. El35-166.The real rate is constructed as the nominal rate minus inflation. Historical Statistics of the United Slates, US Depa-tment of Commerce.
The role of other factors, from budget deficits to the New Deal the set ot programs put in place by the Roosevelt administration to pull the US cconomy out ol thc Great Depression—is less clear.
One New Deal program was aimed at improving thc functioning ol banks by creating the Federal Deposit Insurance Corporation (FD1C) to insure checkable deposits and to avoid bank runs and bank failures. And, indeed, there were few bank failures after 1933.
Other programs included reliel and public works programs lor the unemployed, and a program administered by thc National Recovery Administration NRA to establish 'orderly competition' in industry. Economists generally agree that these programs had few direct cllccts on the recovery. But some economists argue lhat the indirect cflccts ol these programs—particularly the perception ol ihe government's commitment to getting thc economy out ol ihe depression—were important in changing expectations in 1933 and alter. We saw in earlier chapters how such expectational effects of policy can be important. However, showing their importance in 1933 and alter is difficult and remains largely ю be done.
Thc recovery also presents us with a puzzle. In 1933, dellation stopped. The rest ol ihe decade was characterised by small bui positive inflation. The CPI was 81.8 in 1940, compared with 75.6 in 1933. Thc end ol deflation probably helped thc recovery. Thc shift from deflation to rough price stability implied much lower real interest rates than had been the ease trom 1929 until 19.33.
Thc puzzle is why deflation ended in 1933. With a large deflation in 1932 and unemployment ai an all-time high, ihe theory ol wage determination developed in previous chapters implies thai there should have been further large wage cuts and lurthcr deflation. This isn't what happened. As we saw in thc Phillips curve diagram constructed lor Australia (Figure 8.11 the years 193.3-39 arc clear outliers and exactly the same could be said lor the United States. So, why did dellation stop?
• One proximate cause may be thc set ol measures taken by the Roosevelt administration. The National Industrial Recovery Act {NIRAl, signed in June 193 3, asked industries to establish minimum wages, and not lo take advantage ol the high unemployment raic to impose further wage cuts on workers. Economists are usually doubtful lhat such admonitions to firms have much effect. Bui the NIRA offered lirms a carrot in exchange in effect, a decrease in competition in goods markets under the guise ol orderly competition', and thus thc potential lor higher profits il they complied. Thc evidence suggests that the NIRA did have an effect on wage setting.
• Another lacior may be that although unemployment was still high, output growth was high as wel As a result, there were bottlenecks in production, leading firms to increase their prices given wages. Because of thc sharp increase in demand, the price of raw materials was also bid up, increasing costs and again forcing lirms to increase their prices given wages. In short and in contrast to our simple specification ol pricc setting where we assumed that prices depended only on wages, the effect of fast growth was to increase prices given wages, thereby reducing thc deflationary pressure of unemployment.
• A relevant fact in this discussion is that dellation ended in the mid- 1930s in most countries, even in countries lhal didn't have programs similar lo ihe New Deal and didn't have thc same fast growth rates after 1933 as the United States. This suggests that other more general lactors might have been at work. One possibility, which has been explored in thc case ol European unemployment in the 1980s and 1990s. is that, after a while high unemployment exerts less pressure on inllation. Thc idea is that once people have been unemployed lor a long time, thev give up on finding a job. becoming in effect irrelevant to the wage determination process. (We lirst raised this issue in thc discussion on long-term unemployment in Chapter 6. i As a result, unemployment has less elfect on wages, and in turn less ellect on inflation.
Why should we care about how deflation turned to inflation in the United Stales in 1933? Because, as you will sec next, the answer is very relevant to Japan today. How lo get rid of deflation, and, in so doing, decrease the real interest rate and stimulate growth, is one of the main issues confronting Japan today.
Комментариев нет:
Отправить комментарий