[ Pobierz całość w formacie PDF ]
.On the other hand, to theextent that intervention is a policy tool that is used in the context of aloose notion about the rate that is consistent with long-term economicand financial trends, it would be compatible with a framework forexchange market intervention that relied on the coordination channel andsporadic operations for its effects.To summarize, my reading of both the economics literature on theeffectiveness of intervention and my assessment of the actual use of theinstrument by G-3 authorities in recent years is essentially the same as itwas a decade ago (Truman 1994, 249): the evidence is sufficient to supportthe continued judicious use of intervention as a supplementary policyinstrument. 12 Even Sarno and Taylor (2001, 862), who, as noted earlier,are in the camp of those who are positive about intervention s effective-ness, state that the studies allow us to conclude cautiously that officialintervention can be effective, especially if it is publicly announced and10.Many consider the August 15, 1995, joint Japan-US operation as a classic example ofsuch opportunism, going with the flow of the market (Bank for International Settlements1996, 101).11.This term, drawn from American literature, was often used by Sam Y.Cross, a formerUS Treasury official, who was manager of foreign operations for the Federal Reserve SystemOpen Market Account from the early 1980s through 1991, to warn against the risk of enteringthe market without an exit strategy.12.To emphasize here a point that has applied throughout this paper, the issue is interven-tion involving the G-3 currencies.Exchange rates involving less international currenciesmay be more responsive to intervention, or may be responsive for a longer run, becauseof either capital controls or other aspects that make them much less perfect substitutes.254 DOLLAR OVERVALUATION AND THE WORLD ECONOMY concerted and provided that it is consistent with the underlying stanceof monetary policy. 13Two aspects of this statement deserve emphasis: First, they concludethat intervention can be effective, which is not the same as saying thatit is always effective.Second, they lay down three conditions in which itis more likely to be effective: public announcement, multilateral engage-ment, and policy consistency.14 While public announcement is simple andnow common practice among the G-3 authorities, the other two conditionsare more demanding.They are discussed further in the second part ofthis paper.Reaching international agreement that now is the time tooperate in the foreign exchange market is a tedious process, in part becausethe interests of two or more sets of authorities may differ, and in partbecause their views may differ on the effectiveness of the instrument andthe costs of its overuse.Moreover, frequently there is a lack of consensusthat an intervention operation would be consistent with underlying mac-roeconomic policies.One consequence is that attempts to establish guide-lines for G-3 intervention operations such as in the 1987 Louvre Accordare destined to fail within a few days or weeks as soon as conditions andattitudes change to destroy the consensus.Exchange Market Intervention:Policy ConsiderationsFrom a US or national perspective, the overriding objective of macroeco-nomic policies is to achieve maximum sustainable noninflationary growth.In this context, the foreign exchange value of the dollar and the US currentaccount and international investment position are not policy objectives.Those variables also do not systematically affect the achievement of theprimary policy objective using the instruments of monetary and fiscalpolicy.Policymakers, reflecting the views of the general public, may havepreferences about the allocation of fully employed resources betweensectors producing traded (manufactured) goods and services and sectorsproducing nontraded (primarily nonmanufactured) goods and services.They also may have concerns about the sustainability of the US currentaccount balance or international investment position.However, the evi-13.Ito (2002) as well as other researchers have interpreted their results as consistent withthe view that coordinated intervention is more effective, or more likely to be effective.Galati, Melick, and Micu (2002) test this proposition directly and find that coordinatedoperations do not add to the significance of the intervention.14.Dominguez and Frankel (1993) also recommended that intervention should be unantici-pated, publicly announced, and coordinated.They also argued, as Dominguez recounts inher conference paper, that intervention was least likely to be effective if it was inconsistentwith either future monetary policy intentions or future exchange rate fundamentals. THE LIMITS OF EXCHANGE MARKET INTERVENTION 255 Figure 12.1 US output gap, monetary policy stance, and realexchange rate index, 1981-2002Sources: Output gap: OECD Economic Outlook, Nos.65, 67, and 71; real broad dollar index:Federal Reserve Board statistics; inflation data (to compute real federal funds rate): Bureauof Labor Statistics.dence from empirical studies, as discussed above, is that they lack aninstrument independent of the settings of monetary and fiscal policiesto achieve, with any reliability or consistency, the desired allocation ofproduction across sectors or to alter the external accounts.Policymakers, of course, do take account of actual and potential devel-opments in exchange rates and external accounts when making policyand balancing risks.For example, they try to anticipate the effects ofexchange rate depreciation on the real economy and thereby on inflation,they try to anticipate the tendency of exchange rate appreciation todampen the real economy, and generally they are alert to the possibilitythat an unsustainable position in the external accounts will eventually becorrected.That amounts to good analysis, but it is not the same thing asdirecting economic policy at an exchange rate target or at the currentaccount
[ Pobierz całość w formacie PDF ]