sign in sign up
<<
Home , Central bank

What Can Central Do About the of the Dollar?

Dallas S. Batten and Mack Ott

VER the past four years, the stubborn strength of authorities of the and its major trading the U.S. dollar in foreign markets has created partners should buy and sell foreign . The what many observers believe is a precar’ious balance first set of critics urges this policy to manage the fall of between two undesirable features of the floating ex- the dollar’s foreign exchange value, while the second change rate system. On the one hand, some argue that urges it to bring down the value of the dollar in order to 2 the dollar’ is overvalued and, consequently, soon will stimulate the U.S. traded sectors. fall dramatically, generating increased U.S. The purpose of this paper is to explain the fun- and jeopardizing world financial stability. On the other damentals of cents-al intervention in foreign ex- hand, other-s asgue that the “high” dollar was a major change mam-kets and the conditions required for it to be cause of the 1981—82 economic and now effective. First, the motives, mechanics and conse- assert that the cur-rent expansion is not sustainable quences of intervention are discussed. Next, the rela- without a substantial decline in the foreign exchange tionship between intervention and domestic monetary value of the dollar.’ policy is investigated. Finally, some qualitative and Surpsisingly, both sets of dollar critics have ad- quantitative evidence on the efficacy of intervention is vanced the same policy prescription: intervene in fos- reviewed. eign exchange mas-kets. In other- words, the monetary WHY CENTRAL BANKS INTERVENE

Dallas S. Batten and Mack Ott aresenior atthe Federal The is the of one country’s cur- Reserve Bank of St. Louis. Paul G. Christopher provided research mency in terms of another. As the relative price of two assistance. assets (currencies), it is determined by the fos-ces of ‘For examples, consider Kraus (1984), Arenson (1984) and Bank for International Settlements (1984). In the last (p. 1), Dr. Gleske, a demand and supply, as are the of other assets, member of the Directorate of the , was some- what bemused by the dollar’s persistent high value: There therefore seem to be enough reasons to support expectations ‘Traded goods are those goods that are potentially exportable or that the dollar rate will decline. However, contrary to all historical experi’ importable — whether or not they are actually oonsumed domestical- ence, these expectations have not been fulfilled as yet. The markets do ly or abroad. For instance, agricultural commodities, airplanes and not seem to have any expectations of this nature, either. Apparently, steel are tradedgoods, while haircuts, legal services and housing are investing in dollar assets — especially in the united States itself — primarily nontraded goods. The importance of the distinction be- continues to be lucrative despite the greaterexchange rate , and the tween traded and nontraded goods is that changes in foreign com- markets appear to now to have hardly any doubts that the dollar will remain attractive despite the tact that the fundamental factors like the petition will directly affect production and sates both in the home and declining competitiveness of U.S. industry and sharply rising current- foreign markets in the traded goods sector, while only indirectly account deficits operate rather in Savour ot expecting the opposite. affecting produotion and sales in the nontraded goods sector.

16 BANK OF ST. LOUIS MAY 1984

3 such as stocks, bonds or real estate. Moreover, the intermediate term can be more than just a nuisance to relative valuations of and yields on non-currency monetary authorities—they also can have real effects. assets also have, as we shall see, large impacts on the For example, an appreciation of the exchange rate exchange rate. beyond that necessary to offset the inflation differential between two countries (termed “overshooting”) raises Unlike the prices of services or nondurable goods, the price of traded goods in the home country relative asset prices reflect primarily the ’s expectations to the prices of traded goods in the rest of the world. about future economic conditions. Consequently, in Thus, home country exports become less competitive the run, exchange rates should be influenced in world markets and home country import substitutes predominantly by new information — that is, surprises less competitive in domestic markets. Consequently, — which alters expectations of future events; these sales of traded goods decline, generating unemploy- surprises lead to highly unpredictable and often siz- ment in the traded goods sector and, subsequently, able movements in exchange rates. Because news can inducing a shift of resources from the traded to the be incorrect and because markets can oyerreact to nontraded goods sector. news even when it is correct, monetary authorities typically believe that much of the short-run volatility This reallocation of resources is efficient for the 4 exhibited by foreign exchange markets is excessive. economy ifthat portion of the exchange rate apprecia- Consequently, intervention is frequently rationalized tion in excess of the inflation differential is permanent. by central banks as a means to reduce the presumed Ifthis excess portion of the exchange rate appreciation excessive variability of exchange rate movements re- is short-lived i.e., reversed in the near futurel, the 5 sulting from the variability of market expectations. corresponding movement of resources will be re- versed, and the economy will have experienced un- In the run, movements of exchange rates tend necessary due to the costs of shifting toward a relationship among currencies known as pur- resources, reallocating , laying off and hiring. chasing power parity (PPP). That is, a dollar or yen or Monetary authorities who wish to avoid such situa- deutsche mar-k 1DM) would each purchase the same tions may intervene to oppose exchange rate move- amount of goods whether expended at home or ments that they believe will riot persist. This is a fea- abroad. Thus, a country with a relatively rapid inflation sible policy, however, oni if they can distinguish rate will have its currency decline in value relative to temporary exchange rate fluctuations from permanent the currencies of countries with slower inflation rates. ones.° Ignoring long-run considerations, frequent and Exchange rate changes also affect the general price offsetting exchange rate movements in the shor’t or level and may gener-ate some measured short-run in- flation or- as markets adjust to the changing relative price of traded to nontraded goods. In particu- lar, an exchange s-ate depreciation raises the domestic ‘Policymakers sometimes argue that it is too important to be deter- mined solely by market forces. Consider, for example, the view of Dr. currency pt-ice of imports and, thus, raises the domes- Otmar Emminger, former President of the Deutsche Bundesbank, as tic .Because the total impact of this change is expressed in a speech to a Bankers’ Forum at Georgetown Universi- ty, Washington, D.C., on September 28, 1981 (see Bank for Interna- not felt all at once, the price level continues to rise for tional Settlements [1982], p. 4): some time. Thus, since exchange rate depreciation usually precedes changes in domestic prices, it may The exchange rate is tar too important a price — because of its effects appear to cause inflation.’ on domestic stability and activity, the , and interna- tional and payments relations —to betreated with ‘benign neglect’ as a ‘residual outcome’ of domestic monetary and other economic policies. This should not be misunderstood as a call for activism and 6 interventionism in the field of exchange rates, It cannot be emphasised This distinction is difficult to make in practice. As Martin Feldstein too much that the most important foundation for exchange rate stability is a policy of monetary and financial stability at home. We would, (1983), p. 48, observed: however, delude ourselves it we believed that all that is required for stable exchange rates is the pursuit of a steady supply policy. [Tlhere is no way in practice to distinguish an exchange-rate movement that is merely a random fluctuation from one that is part of a fundamental Short and medium-term strains and disturbances are built-in elements shift in the equilibrium exchange rate, Exchange-market intervention in the field of payments and exchange markets, even if we had firm aimed at smoothing a transitory disturbance may infact be a counterpro- international agreement and successful implementation of concerted policies. ductive or futile attempt to prevent a basic shift in the equilibrium ex- change rate. 4 This concern, however, appears to be misdirected as Bergstrand Furthermore, this overshooting may 000ur because asset markets (1983)has demonstrated that exchange rates are actually less vari- clear more rapidly than do goods markets, and, as a result, it may not able than other asset prices. be undesirable. 5 For additional supportforthis characterization, see Kubarych (1983), ‘For a more detailed discussion of the relationship between domestic pp. 16—19 and 43—45. inflation and exchange rate changes, see Batten and Ott (1983).

17 OF ST. LOUIS MAY 1984

Finally, many domestic residents, firms and, espe- employed method — intervention by the monetary cially, multinational have financial assets authority. and liabilities denominated in foreign currencies. Ex- change rate changes, then, produce wealth effects it Typical Example since they generate capital losses and gains. For exam- ple, if the U.S. exchange rate unexpectedly appreciates, Suppose that the dollar is believed to be overvalued. the dollar values of foreign-currency-denominated The Federal Reserve Bank of New York, which acts as assets and liabilities fall. Hence, U.S. monetary the for U.S. interven- debtors in foreign currencies experience gains, and net tions, will purchase foreign currency, typically PM, 1 creditors experience losses? with U.S. dollars. ° It can do this simply by creating dollar reserves and using them to purchase PM. In In sum, changes in exchange rates have conse- particular, the Fed can purchase PM-denominated de- quences that monetary authorities may deem unde- posits of U.S. banks at German banks and pay for them sirable. Thus, having chosen not to allow exchange by crediting the reserve accounts of these U.S. banks. rates to be completely market-determined, many cen- The Fed then presents to the Bundesbank drafts drawn tral banks intervene periodically in foreign exchange against accounts of these U.S. banks at German banks, markets to mitigate what they believe to be transient which are subsequently cleared by the Bundesbank. but deleterious effects of exchange rate movements on The impact of this transaction on the financial institu- the domestic economy. tions involved is outlined in figure 1. In gener-al, the reserves of the U.S. banking system increase, while HOW CENTRAL BANKS INTERVENE those of the German banking system fall. The changes in the reserve positions of the United States and Ger- The mechanics of intervention in for- many that result from this foreign exchange operation eign exchange markets can take a variety of forms. The will cause the U.S. money stock to rise and Germany’s general purpose of each variant, however, is basically money stock to fall. the same: augment the market demand for one curren- cy by augmenting the market supply of another. An Conversely, if the Bundesbank believes the PM to exhaustive explanation of the ways in which interven- be undervalued (i.e., the dollar- is overvalued), it could tion can be conducted is beyond the scope of this reduce the quantity of PM relative to dollars. This paper? Instead, we will describe the most ftequently transaction is slightly more complicated than when the Fed intervenes in support of the PM. First, the Bundesbank must acquir-e dollars. It typically does this eA net monetary foreign currency debtor is an individual or firm with either by selling some of its non-negotiable U.S. Trea- greater monetary than monetary claims in a foreign currency. sury securities to the Fed or by borm-owing from the Thus, when that foreign currency depreciates, his foreign liabilities decline in value more than his foreign assets, and, on net, his wealth rises. Similarly, a foreign currency depreciation would lower the wealth of a net monetary foreign currency creditor. ‘°Inthe United States, the Federal Reserve Bank of New York inter- °SeeBaibach (1978) for a detailed analysis of various forms of in- venes for the Federal Reserve System and the U.S. . The tervention. decision to intervene, however, is made by the U.S. Treasury.

18 FEDERAL RESERVE BANK OF ST. LOUIS MAY 1984

Fed in exchange for a dollar-denominated account ities do not want their foreign exchange market in- through a swap arrangement already established be- tervention to affect their domestic economies, they tween the two. These acquired dollars are then used to may sterilize its impact with an offsetting sale or buy PM in the foreign exchange market. purchase of domestic assets In figure 2, step 1 depicts the acquisition by the Sterilized intervention would be the preferred pro- Bundesbank of a dollar--denominated at the cedure if the Fed did not want the U.S. banking sys- Fed. Since this tr-ansaction is between central banks, it tem’s reserves to change. Thus, if the unsterilized does not affect the reserves of the banking system in intervention interfered with the goals of domestic either country and, hence, does not affect either- coun- , the Fed could sell US. Treasury tsy’s domestic money supply. In step 2, the Bund- securities in U.S. financial markets equal to the amount esbank purchases PM-denominated deposits of U.S. of reserves created by the intervention. With this trans- commercial banks held at German commercial banks action, the level of reserves in the U.S. banking system with dollars. This transaction is cleared by U.S. banks would return to its preintervention level, and, as a presenting to the Fed dollar-denominated claims result, there would be no subsequent change in the against the Bundesbank and receiving reserves in re- U.S. money supply. turn. (At the same time, the Fed reduces its deposit Similarly, the Bundesbank could neutralize the im- liabilities to the Bundesbank.) Likewise, the Bund- pact of intervention on the German money supply by esbank clear’s the draft it purchased from U.S. banks by injecting new reserves into its banking system. Ifsteril- lowering its reserve liabilities to Get-man banks. And ized completely, the foreign exchange operation finally, German banks, presented with a draft against would not affect either country’s money supply. Thus, deposits of U.S. banks, r-educe their deposit liabilities to in the case of a completely sterilized intervention, these banks by the amount of the reduction in their private portfolios would contain fewer PM- reserve deposits at the Bundesbank. The final result is denominated securities and more dollar-denominated the same as in the preceding case — the r-eserves of the securities, while the Fed’s portfolio would contain U.S. banking system rise, while those in the German more PM-denominated securities and fewer dollar- banking system fall. denominated ones.

Sterilized vs~Unsterilized Intervention THE EXCHANGE RATE The two examples discussed above are instances of CONSEQUENCES OF CENTRAL unsterilized intervention; that is, the domestic money BANK INTERVENTION supplies have not been insulated from the foreign ex- change market transaction. Ifunsterilized intervention Exchange rate movements reflect two fundamental is undertaken in large amounts, it will affect not only characteristics that must be recognized to understand the money supplies of both countries, but domestic the impact of intervention. First, changes in exchange prices and rates as well. If monetary author- rates, like changes in the price of any asset, are highly

19 FEDERAL RESERVE BANK OF ST. LOUIS MAY 1984 irregular and unpredictable in the short run, reflecting asset prices unless their activity is substantial or their primarily new events that alter the market’s expecta- actions affect market expectations. tions of future exchange rate movements. Second, This intervention may affect market expectations since the exchange rate is the relative price of two about relative asset yields and prices if market partici- currencies, its movements reflect changes in policies pants interpret the expansion of the U.S. money supply that affect either the supplies ofthese currencies or the as an indicator of a permanent increase in the rate of demands to hold them. More specifically, exchange monetary growth planned by the Fed. Such expecta- rates reflect anticipated relative inflation rates that are tions of further easing of U.S. monetary policy will generated by both past and expected future policy cause market participants to anticipate increases in actions of the countries involved.Therefore, currencies the rate of U.S. inflation relative to German inflation. of countries with lower expected inflation rates are These expectations ofrelatively higher future U.S. infla- cheaper to hold and are in greater demand than those tion will decrease the desire of foreigners to hold dol- of countries with higher inflation rates, all other things lars since they expect the dollar’s purchasing power to equal. Consequently, it follows from PIt that curren- continue to fall. Consequently, the PM value of the cies of higher-inflation countries tend to depreciate dollar will depreciate at a rate equal to the difference relative to those of lower-inflation countries. between the now higher anticipated rate of inflation in 11 The different impacts of sterilized and unsterilized the United States and that in Germany. intervention on the exchange rate can be analyzed in Ifthe intervention is sterilized, its immediate impact terms of these two characteristics. Suppose, as out- is the same as that for unsterilized intervention; that is, lined above, that either the Fed or the Bundesbank it generates a temporary increase in the flow demand purchases PM with U.S. dollars. This transaction in- for PM. The net effect of this sterilized intervention is creases the short-run flow demand for PM relative to simply that private portfolios will contain more dollar- the supply of PM and should result in an appreciation denominated and fewer PM-denominated securities; of the PM, all other things equal. This impact will be neither country’s money supply will be affected. Con- only transitory, however, unless the central bank con- sequently, it is not clear what lasting impact this type of tinues to purchase PM day afterday, ther-eby maintain- intervention will have on the PM/dollar exchange rate. ing the higher flow demand for PM. Because sterilized intervention entails asubstitution More importantly, since this tr-ansaction is unsteril- of dollar-denominated securities for PM-denominated ized, it causes the U.S. money supply to rise and the ones, however-, the exchange rate will be permanently German money supply to fall. If large enough, this affected only if the investors view domestic and foreign intervention has two potential effects: a transient effect securities as being imperfect substitutes. If this is the on the current markets for the two currencies and a case, investors will be unwilling to hold the new port- permanent effect on expected future relative inflation folio at unchanged exchange and interest rates. In fact, rates. Other things equal, the resulting excess supply of at the original exchange and interest rates, an excess U.S. money in the United States and excess demand for demand for PM-denominated securities will arise. German money in Germany will cause the two coun- Consequently, investors will attempt to acquire addi- tries’ money markets to clear at lower and higher-rates tional PM-denominated securities in order- to retur-n ofinterest, respectively. This immediate, but transitory, their portfolios to the desired proportion of dollar’- effect will cause the dollar to decline relative to the PM denominated securities, ther-eby placing downward as German assets temporarily have higher yields than pressure on the PM value of the dollar. U.S. assets. If investors consider these securities to be perfect substitutes, on the other hand, no change in either the Whether this effect on the exchange rate is lasting or transient, however, depends crucially on the expecta- tions of investors holding U.S. and German assets. If these expectations are unchanged, individuals holding “This discussion is oversimplified in that it isolates only two curren-

U.S. assets would sell them — driving up theiryields — cies and the exchange rate between them. In the real world, there are numerous currencies and exchange rates. Attempts to affect the and buy German assets — depressing their yields — exchange rate between any pair of currencies necessarily affect not thereby tending to offset the central bank’s actions. only the exchange rate between this pair, but all other exchange That is, just as a single private individual in a competi- rates as well. Consequently, intervention to move one exchange rate in a desirable direction or to calm fluctuations in that exchange tive market can have no effect on asset prices by his rate may cause another exchange rate to move in an undesirable sales or purchases, so even central banks will not affect direction or to become more volatile.

20 FEDERAL RESERVE BANK OF ST. LOUIS MAY 1984 exchange rate or in interest rates will be necessary to change rate management from money stock control, motivate investors to hold this portfolio. In summary, however, depends on whether certain conditions are when two domestic money supplies have been un- met. First, international assets (including currencies) affected by an intervention activity, the intervention must be imperfect substitutes. Second, the magnitude can have a permanent impact on the exchange rate of sterilized intervention undertaken must be large only if foreign and domestic securities are imperfect enough — given the degree of imperfect substitutabil- substitutes. ity — that market participants cannot undo this effect by engaging in offsetting transactions.

INTERVENTION AND DOMESTIC SOME EVIDENCE ON THE MONETARY POLICY EFFECTIVENESS OF INTERVENTION

The foregoing discussion has emphasized that the Assessing the efficacy of intervention is difficult be- relationship between domestic monetary policy and cause data on central bank intervention are not made intervention depends on whether the intervention is available; in contr-ast to domestic central bank transac- sterilized. Pomestic monetary policy cannot be con- tions, which are reported in great detail, international ducted independently of unsterilized intervention transactions are reported only in a non-systematic, since, as discussed above, it is tantamount to conduct- summary form. Three pieces of qualitative evidence, ing monetary policy through foreign exchange market however, can be used to gauge the likely effectiveness operations. Thus, the exchange rate is a third alterna- of intervention. The first is an indirect assessment tive monetary target variable to those more frequently obtained by considering adomestic policy experiment, considered — namely, monetary aggregates or interest somewhat analogous to sterilized intervention, which rates. occurred in the early 1960s. The second is an assess- Because there can be only one monetary policy ment of the potential for the U.S. tnonetary authorities stance, the role of unsterilized intervention depends to influence the foreign exchange market by compar- crucially on the importance that policymakers place ing the volume of assets and the rate of transactions in on the exchange rate as an objective for monetary these markets by private investor’s with the monetary policy. In particular, the use of unsterilized interven- authority’s holdings and activities. The third is a direct tion necessarily implies that the assessment of U.S. and other central bank intervention places relatively more importance on reducing the activity revealed in a working group study prepared for and real economic disturbances associated with the 1983 Williamsburg Economic Summit Meeting. exchange rate movements than achieving domestic tar-gets for- inflation and unemployment. The manip- An Analogous Policy: Operalion Twist ulation of monetary policy to achieve exchange rate A historical example of a domestic policy experi- objectives inevitably will conflict — occasionally or ment by the Federal Reserve that is similar to sterilized frequently — with the policy stance required to intervention is Operation Twist.” Puring 1961-432, the achieve these domestic objectives. Federal Reserve sold short-term U.S. securities and Fur-thermore, exchange rate movements may be used the proceeds to buy long-term U.S. securities; as motivated not only by changes in the desire to hold with sterilized intervention, the transactions were domestic currency which probably should be offset offsetting so that the money supply was unchanged. by changes in the domestic money supply), but also by ‘rhe resulting increase in short-term gover-nment a host of other factors, especially the policies followed securities and the concomitant decline in long-term by foreign policymakers. Pirecting domestic monetary government securities in private portfolios were in- policy at an excharlge rate target, then, subjects the tended to raise the on short-term securities and lower the yield on long-term securities, thus, twist- domestic economy to disturbances from both domes- 12 tic and foreign sources. Consequently, the monetary ing” the term structure. authority loses its ability to control its own money In this effort, the first condition discussed above was supply independently of foreign events. met — namely, long-term securities bear higher- in- The desire to influence exchange r-ate movements terest rates than short-term securities, and, thus, the wrthout losing control of the domestic money supply is the primary motivation for using sterilized interven- 2 tion. Whether a monetary authority can separ-ate cx- ‘ See Markiel (1966), pp. 219—43.

21 FEDERAL RESERVE BANK OF ST. LOUIS MAY 1984

Chart] Relative Stocks of Holdings of Foreign Assets Billions of dollars Billions of doltors 800

700 Private U.S. nondomeslic asse(s Li. —- 100 Foreign-owned assets in the United Stales LI.

600 U.S. total foreign reserves minus L~ 600

500

400 400

300

200 200

100 100

0 —Is 0 1913 1914 1975 1916 1977 1918 1979 1980 1981 1982 Saurces, Board af Gaverears f the Federat Reserve System and u.s. Department al Commerce LI Year-end outstanding. 0 ~ Annual averages at quarterly data. NOTE, Direct data priar to 977 do nat reflect t977 benchmark revisian. two assets are imperfect substitutes. Yet, the effort is lion of currencies and securities denominated in those generally judged to have failed primarily because the currencies is altered; two market activities, however-, 13 policy was not executed vigorously enough. The are thereby set in motion that tend to undo any point (as emphasized by Malkiel) is that a central bank impact on relative interest rates and the exchange rate. policy of affecting the term structure for interest rates First, private entities — banks, primarily, but also indi- depends, for its effectiveness, on two points: the vidual traders — sell or buy securities denominated in bank’s ability to affect significantly the r’elative supplies the currencies that have been affected. Second, actual ofshort- and long-term financial assets and its willing- cun-ency flows and options to buy or sell currencies or ness to do so. In this case of intervention in domestic forward contracts are changed. Thus, unless the cen- asset markets, the extent of the activity was inadequate tral bank is prepared to take sufficient actions to alter to have any significant impact. tnarket expectations, it will be unlikely to affect the exchange rates by sterilized intervention. Similarly, when sterilized intervention in foreign ex- change markets is undertaken, the immediate distribu- The Potential: Foreign Asset Holdings 13 As Johnson (1963), cited in Malkiel, p. 234, concludes: and the Size of the U.S~Foreign Whatever might have been expected of this policy. ..it was not in fact Exchange Market pursued in any effecrive sense. As a result primarily of Treasury fund- ing operations, the maturiry of the debr in r he hands of the public has in The likelihood that intervention can affect exchange fact been lengthened appreciably, instead of shorlened as the policy would require. rates may be assessed by comparing either private and

22 FEDERAL RESERVE BANK OF ST. LOUIS MAY 1984

Chart 2 Relative Investment Flows into Foreign Assets

Billions of dollors Billions of dollars 120 120

100 100

80 -______.. / ~ 80 Changes in private U.S. asset holdings abroad a 60

s ~:j ~ge~d~ 40

20 hoidings in the United Stalesu, 20

~ ~ ~ - -~ Ne-’

0 -,- —------— 0

U.S. loreign exchange ~-- intervention t2 -20 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 Sources, Bawd of Gavet-nars of the federal Reserve System and u.s. Deportment of Commerce U, Annual rate. [2Pasitive values are purchases of dallars. NOTE, Direct investment data priar to 1977 do not reflect 1977 benchmark revision.

central bank for’eign asset holdings or central bank lar-ger share of assets traded in international markets activity and the size of markets for foreign exchange. than do the Fed and the U.S. Treasury. Consequently, The relevance offoreign asset holdings is that changing to change the pnce at which these assets are valued any single exchange t-ate changes the price of these would require very aggressive intervention. assets in two or’ more currencies. Consequently, ex- change rate movements caused by intervention may be Many have argued that the primary impact of in- presumed to induce shifts in desired portfolios of tervention is on the flow demands for the currencies involved. If so, one should compare the flows of trans- assets — that is, flows of asset sales and purchases t4 that tend to offset such changes. actions in these markets rather- than asset holdings. From this perspective, consider chart 2—transactions Chart I shows the stocks of foreign assets held by in international markets. Comparing the rate of ac- U.S. individuals and institutions, U.S. assets held by private foreigners, and the foreign reserves (minus 4 gold) of the U.S. Federal Reserve System and the Trea- ‘ This leaves aside, for the moment, the indirect effect of intervention through changes in expectations offuture central bank policy, which sury. It is clear that private investors hold a much will be considered later.

23 FEDERAL RESERVE BANK OF ST. LOUIS MAY 1984

\,~ :~c,~J ,~s, ~~t~L‘~,

, I 4S , ~ L~’ ~ I~\ ~ ~Lcy <‘~ ‘ ~L/‘ ,~i\‘~‘f”~ ~~~ ~ ttlp~~er ‘IL ~ “ ~ .)~~:: ~—:~1& ~ ~~Møø ~ c~~

~ ~r~c:~ ‘5 ~s ~ ~ ~Jwk~w. ~A$fl~ØL/ > ~ ~~~‘S ,, L ,,,/cLv~ ~H ~ ~&, /s.~‘

Ii ,,sc /r ~‘ ~ ~ — /4 “ ~ c~ ‘~ ~ ~,, ~‘ 7 <‘~S,Pz ~~ ~ ./ — ~4k ~~ ,, ~/,4. ~ ,.5~ C / ,, ~ c.~ ~N C~z .~, L <~ ~‘ & ~s, ,~.‘ ~ k .~ 4 ,,~Lt ~ ~ ‘s/ ~ “ ‘/ ,~ ~ ~‘ .~ ~4~\ ~ ~ ~ ~, ~, I 4’~~ ‘~‘ r “ ~ ~ ‘7 V C’. ~ k..V.’ ,.t , L’

~~:rt~~t Tff%~flV1~i~trw ~~~ V .7,. ~ ~. ~ ~. ,\ V ~ ,~ &V •.u~aL V ~ ‘S~<” ~ ,.~ :1 ; ~ ~ .‘ : MV ~ N /V ~ V V1~lS~$o: ~j~:EIV~ ~rnNItm~~~ V~ ~ i

N~:: ~ ~NV :N :,:~::~~ ‘~ ~ VN> V~1t~V~ ~: !t~L ~ ~,N

quisition offoreign assets by U.S. investors with the rate for U.S. banks engaging in significant volumes of foreign 55 ofintervention by U.S. monetary authorities (combined currency transactions. As the table shows, both the U.S. Treasury and Federal Reserve) clearly demon- volume in the observed month) per bank and the strates that the private U.S. investment activity on number of banks with significant involvement in for- an annual basis swamps that of the authorities, and, eign currency markets have risen dramatically since in fact, so has the rate of purchase of U.S. assets by 1977. The total volume has risen sevenfold, comprising foreigners in recent years. a quadrupling in the per-bank volume indicated by the changes in the activity of the banks originally sur— Another, perhaps even more relevant. comparison veyed in 1977) and a tripling from 44 to 119) of the would be the transaction rates over intervals shorter number of banks actively participating in foreign cur- than a year. That is, it might be argued that chart 2 shows net figures over an irrelevantly long time period: t what matters is the gross volume of transactions in, ‘ The data in the first half of table 1 are from periodic surveys of U.S. banks that engage in significant foreign exchange market transac- say, a month or a day. From this perspective, consider tions. These surveys are conducted by the Federal Reserve Bank of the data in table 1, which r-epor-ts the turnover statistics New York. For more details about these surveys, see Revey (1981).

24 FEDERAL RESERVE BANK OF ST. LOUIS MAY 1984 rency markets. If the volume in contracts during April domestic policy, especially monetary policy. When in- 1983 is expressed in a daily aver-age form, the market in consistencies have arisen between domestic policy U.S. banks alone is about $33.5 billion. Almost two- and exchange rate objectives, the group found that thirds of these contracts are spot currency exchanges. intervention (counter to the goals of domestic policy) was frequently useless and even counterproductive in In comparison, the table also reports the combined the absence of supportive domestic policy. Conse- Federal Reserve-U.S. Treasury intervention for the full quently, the ministers, in their statement released with six-month period containing each of the three survey the working group’s report, downplayed the impor- dates along with representative episodes of significant tance of sterilized intervention as a separate policy US. intervention. As even a cursory review of the data tool: reveals, U.S. intervention activity has been trivial rela- tive to the volume of bank trading in currencies; only We have reached agreement [that), under the presenr circumstances, the role of [sterilizedl intervention can rarely has intervention been morn than a tiny fraction only be limited. -. - Intervention will normally be useful of the private market volume. Consequently, the notion only when complementing and supporting other that the central bank has influenced the market price policies 58 of currencies — their exchange rate — purely by affect- ing the flow volume of exchange is inconsistent with CONCLUSION the recent record. Most discussions of the effectiveness of central bank Some Direct Evidence: Report of the intervention focus on expectations of market partici- Working Group on Exchange pants and how intervention alters them. Yet, even if the central bank is capable of altering market expectations Market Intervention about its future policies, such a change can be brought about only if the market is convinced that other policy Central bank intervention — whether in domestic asset markets or international currency markets—can goals — the domestic inflation rate, level of interest rates, stability of domestic markets, etc. — are be effective only if the market is convinced that the subordinate to exchange rate manipulation. For the monetary authority is both able and willing to affect the United States, at least, such a policy stance would not flows of transactions. In view of the growing size of be credible. Thus, the efficacy of exchange rate in- markets and the conflict with domes- tervention would seem to be diminished greatly by the tic inflation policies that effective intervention would public’s knowledge of the primacy of other monetary require, such an effect on market expectations also 19 seems to be beyond the grasp of the U.S. authorities. policy objectives.

Support for this conclusion is provided by the study The model’s verdict was that the Bundesbank has little if any power to influence the exchange rare over that lime span [one month] without of exchange market intervention conducted by the altering currentor expected future money-market conditions [i.e., with- working group established at the Versailles Summit in out conducting unsterilized intervention]. I sz.~~This report is especially significant since it ltStatement on the Report of the Working Group on Exchange Market 9 represents the most comprehensive analysis of the Intervention (1983), p. 2. 0 motives, methods and impacts of central bank inter- ‘ Consider again the view of former Bundesbank President Em- vention in foreign exchange markets that has been minger (Bank for International Settlements [1982j. pp. 5—8): To sum up: Exchange rare policies cover a wide spectrum: trom simply conducted using actual intervention data — data un- ‘having a view’ on the exchange rate to smoothing out disorderly available to most researcher-s. conditions’ to avoiding excesses which are palpably far out of line with fundamentals and are disturbing. The instruments range from interven- While the working gi-oup found that sterilized in- tion to policy, general monetary and other economic tervention is not totally ineffective, its effect was much policies, and to official borrowing or lending. 7 There is also a wide spectrum in the use ofsuch policies from country to smaller than that of unsterilized intervention.’ Fur- country. This is partly a question of size. For the United States, there is thermol-e, the group found that intervention could he quite certainly no other solution but free floating. The problems of a more active exchange rate policy are mainly relevant br middle-sized effective in the face of persistent mar-ket pr-essures onl industrial countries, It is therefore natural and understandable that the if it was supported by complementary changes in attitude lowards exchange rate policy differs between the United States and other industrial countries. 6 There is also another important difference, which makes the United ‘ fleport of the Working Group on Exchange Market Intervention States a special case. Other industrial countries usually lake the dollar (1983). What follows is a synopsis of the major conclusionsdrawn in us yardstick and intervene against the dollar. For the United States, it is this report. more difficultto decide against which individual currencies they should 7 measure the value of their currency, and against which to intervene in ‘ Additional support, at least for Germany, is provided by Obstfeld case of need. This is one of the several problems on which the oft- (1983). pp. 184—85, as he concludes that: requested joint and concerted intervention policy would founder.

25 FEDERAL RESERVE BANK OF ST. LOUIS MAY 1984

REFERENCES Johnson, Harry 0. ‘An Overview of Price Levels, Employment, and the Balance of Payments,” The Journal of Business (July 1963), Arenson, Karen W. ‘DollarSeen ator Near Peak,” New York Times, pp. 279—89. February 14, 1984. Kraus, Albert L. “Turn in Dollar’s Value Already May Be at Hand,” Balbach, Anatol B. ‘The Mechanics of Intervention in Exchange New York Journal of Commerce, February 14, 1984. Markets,” this Review (February 1978), pp. 2—7. Kubarych, Roger M. Foreign Exchange Markets in the United Bank for International Settlements. Press Review (February 18, revised edition (Federal Reserve Bank of New York, 1983). 1982), pp. 2—6. Malkiel, Burton Gordon. The Term Structure of Interest Rates (Princeton University Press, 1966). Bank for International Settlements. Press Review (February 7, 1984), pp. 1—5. Obstfeld, Maurice. ‘Exchange Rates, Inflation, and the Sterilization Problem: Germany, 1975—1981 ,“ European Economic Review Batten, Dallas S., and Mack Ott. Five Common Myths About Float- (March/April 1983), pp. 161—89. ing Exchange Rates,” this Review (November 1983), pp. 5—15. Report of the Working Group on Exchange Market Intervention estab- Bergstrand, Jeffrey H. “Is Exchange Rate Volatility Excessive’T’ lished at theVersailles Summit of the Heads of and Govern- New Economic Review (September/October 1983), pp. ment, June 4,5, and 6, 1982 (March 1983). 5—14. Revey, Patricia A. ‘Evolution and Growth of the United States For- Feldstein, Martin. ‘The World Economy Today,” The eign Exchange Market,” Federal Reserve Bank of New York (June 11, 1983), pp. 43—48. Quarterly Review (Autumn 1981), pp. 32—44.

26