Microsoft Word - 2007_01_02.doc
Exchange Rate Regimes and Monetary Policies in Emerging
Markets: A Showdown for Few Theoretical Misconceptions
Marko Malović, Belgrade Banking Academy
KEY WORDS: «Two‐Corner Solution», Crawling/Monitoring Bands, «Dirty Float», Flexible Inflation Tar‐
geting, (Un)Sterilized Interventions, Small Open Economy, Emerging Markets, «BBC» Plus.
UDC: 338.23:336.74 JEL : E52, E61, F31, F41.
ORIGINAL PAPER
ABSTRACT: The purpose of this paper is to challenge couple of dangerous theoretical misconceptions
in open‐economy macro, namely, in respect to desirability or sustainability of available exchange rate re‐
gimes and inflation targeting framework and their mutual compatibility in small open economies with in‐
complete (emerging) markets.
First of all, we dismiss the ruling »two corner solution« as dogma in scientific disguise. Furthermore,
all the benefits of more flexible intermediate regimes (sliding currency bands) as well as empirical support of
their wellbeing have been put forward.
As to the monetary policies, majority of transition countries recognised superiority of inflation target‐
ing over alternative monetary concepts. However, until very recently some emerging market economies
failed to realize the benefits of full‐fledged ‐let alone flexible‐ inflation targeting. In what follows, the article
counters another theoretical dogma: that inflation targeting in emerging market economies must go hand in
hand with fully flexible exchange rate regime.
Having said that, and again contrary to the mainstream literature in the field, paper exposes some se‐
rious weaknesses of the so‐called dirty (or managed) floating as an intermediate regime: in particular, its
potential sub optimality in practice and its hidden incompatibility with widespread inflation targeting
strategies.
Paper concludes by reiterating the inevitability of close relationship between inflation targeting and
exchange rate targeting and hence suggests several possible reaction functions for the monetary authorities
in emerging markets among those already laid out in the related literature.
Introduction
This is a theoretical, yet policy‐oriented essay which tries to isolate and address several is‐
sues that have plagued the quest for optimal monetary and exchange rate policy in the emerging
market countries for too long. Voices, opinions and paper‐titles akin to the view taken in this dis‐
cussion are thankfully gaining in strength and professional appreciation, while still remaining
outnumbered and being considered quite a bit unorthodox. Hence, the flickering point in joining
the band wagon of numerous papers written or being written on monetary and exchange rate pol‐
icy issues.
The rest of the paper1 is organised as follows: Section 2 depicts the still ruling dogmatism of
the «two corner‐solution», nonetheless confronted with the criticism of bipolar view and a look
into reality, eventually reinforced by dismantling of the too rigid understanding of impossible trin‐
ity theorem. Section 3 (re)examines the potential, the rationale and empirical tests for relatively
flexible currency bands of the so‐called BBC breed. Section 4 brings in the notion of inflation tar‐
geting as a monetary strategy in emerging market countries, while stressing specificities and some
of the intellectual wanderings that IT regimes have exhibited/undergone in small open economies.
1 This paper partly draws on research carried at and the hospitality of The Centre for the study of Global Governance of
the LSE, where the author was a visiting fellow under the auspices of the FCO/OSI/LSE Faculty Development in South
East Europe Programme.
2007 ‐ 18 • Economic Analysis®
Section 5 looks more carefully into the shortcomings and imminent incompatibility of the popular
dirty floats in the presence of inflation targeting practices in emerging markets thus far. Subse‐
quently, section 5 extends to overall appraisal of rationale and feasibility for simultaneous ex‐
change rate and inflation targeting.
Rejection of «Two‐corner solution»
In modern open‐economy macro, there is a famous theoretical paradigm of the so‐called
«impossible trinity», which reminds us that a country cannot simultaneously keep complete liber‐
alisation of international capital transactions, independence of monetary policy making and fixed
value of national currency. Ignorance or inability to fully appreciate this theorem has recently
placed many economies with fixed but adjustable exchange rate regimes in the epicentre of devas‐
tating international financial crises [Fischer, 2001]. Mexico, Thailand, Indonesia, South Korea,
Czech Republic, former Yugoslavia, Argentina, India and Turkey are some examples of financial
fiascos caused by unsustainable fixed (but adjustable) exchange rates. After the turbulent experi‐
ence of 1990´s, leading theoreticians followed by emerging market policymakers forcefully rejected
the wide spectrum of intermediate exchange rate arrangements between rigidly fixed and freely
floating extremes. This bipolar view2, preached by several leading macroeconomists in the world3
was further acknowledged by the famous Metzler Commission. Bipolar view, alias two‐corner
solution, suggests that developing and transition countries, faced with fast and furious traits of
financial globalisation, have only two feasible options when it comes to exchange rate manage‐
ment, if they are to shelter themselves from devastating effects of currency crises. Namely, either
superfixed regimes like ‐for instance‐ very hard pegs, currency boards and dollarisation, or fully
flexible exchange rates [Fisher, 2001], [Reinhart‐Reinhart, 2003].
Closely related to bipolar view is Hausmann´s concept of «original sin», which claims that
instability of emerging market currencies and low credibility (high sovereign and political risks)
are eternal road‐block to stabilisation and development efforts of their economies [Hausmann,
2002]. Their inabilities to borrow in domestic currency or clear the reputation of budget deficit‐
monetisers have been seen by Calvo as principal causes of the «fear of floating» in emerging mar‐
kets, which also happens to be a rather superficial reason for apparent popularity of dirty floats in
these countries lately [Calvo‐Reinhart, 2000, 2000*]. In other words, although freely floating ex‐
change rates are making extreme balance of payments crises and severe currency crashes next‐to‐
impossible events (if we abstract from somewhat intrinsic FX volatility), despite the fact that they
enable monetary autonomy and preservation of foreign exchange reserves, every more sizeable
depreciation inevitably brings about a drop in living standard (since the wages/salaries are often
sticky in local currency) and tightens the dollar‐denominated debt noose. That, of course, retro‐
spectively points towards the fixed corner of bipolar extremes. Moreover, international finance in
the real world soon enough came out with the holy grail: managed or dirty float, currency regime
that seems acceptable and sufficiently resemblant to fully flexible regime. This «revelation» fed the
bipolar view even more voraciously, thereby epitomising the notion of the non‐existing, or «the
missing middle» [Goldstein, 2002]. In theory, however, managed or dirty floating assumes that the
2 As a result, number of IMF members who positioned themselves in one of the corners along the exchange rate regimes
axis grew in period 1991‐1999 from 39% to 66% [Goldstein, 2002].
3 Larry Summers, being primus inter pares, then more nuanced Stan Fisher, as well as Jeff Frankel and Sergio Schmukler
(although the last two oddly enough changed their mind somewhere along the road) [see Reinhart‐Reinhart, 2003].
Nonetheless, the first who forged the term »two‐corner solution« were in fact Barry Eichengreen, Maurice Obstfeld and
Ken Rogoff in the mid 1990´s, yet from the view point of EMS members, not emerging markets [Frankel et al., 2001, page
354].
Volume 40 • Spring 2007 • 19
value of national currency is being constantly determined in the forex market, only sporadically
accompanied by central bank interventions in accordance with other urgently pressing legitimate
objectives of macroeconomic policy making. Nevertheless, after scrutinising dirty floats through
empirical tests of actual floating (with a null hypothesis that floating currency regimes arguably
need neither gigantic foreign exchange reserves nor too often interventions in the forex market),
practice has revealed that they are much more akin to dirty (camuflaged) fix than a managed float;
see for instance [Hausmann‐Panizza‐Stein, 1999], [Detken‐Gaspar, 2003]. And indeed, it has to be
said that (super)fixed exchange rate regimes do display certain indisputable advantages. In the
case of both currency board and dollarisation, inflation expectations are minimised, exchange rate
‐and overall macro‐ stability is initially achievable whereas currency risk related costs are almost
non‐existing (if reserves under board are big enough to support broad money rather than M1 or
monetary base only). On the other hand, both dollarisation and as currency boards completely tie
down hands of monetary authorities and expose the economy to often asymmetric exogenous
shocks of the business cycle. Furthermore, sooner or later they entail overvaluation risk and immi‐
nent wages‐productivity conflict (for currency boards), or loss of seignorage, legal problems with
outright introduction of foreign currency and inflation typically higher than in the euro/dollar
area as contradictio in adjecto sui generis (for dollarisation).
In a nutshell, there are minimum three crucial reasons why emerging market countries ide‐
ally should adopt neither of the extreme corner solutions, but rather opt for limited flexibily re‐
gimes of the «BBC» type:
• Even a partial autonomy in anticyclical monetary policy making always means a lot for
fragile economies in transition.
• Modern forex and financial markets (and especially emerging ones) are ages away from
stylised Arrow‐Debreu world, therefore much of speculative trading is not stabilising as ones
claimed by M. Friedman, but random at its very best!
• In presence of more serious import dependency and/or significant foreign indebtedness, it
is neither credible nor feasible to expect from monetary authorities to completely give up exchange
rate targeting.
In the context of «impossible trinity» dogma, the argument goes that although we cannot
simultaneously fully liberalise capital flows and retain absolutely fixed exchange rate as well as
absolutely independent monetary policy, nobody have ever said one couldn’t partially control
capital flows, and retain reasonably flexible currency rate and relative independence in formulat‐
ing monetary policy [Williamson, 1999], [Frenkel et alia, 2001]! This impression is amplified by
findings of several recent studies which have confronted de facto exercised with de iure adopted
exchange rate regimes among the IMF members [Reinhart‐Rogoff, 2004], [Frankel et alia, 2001],
[Hausmann‐Panizza‐Stein, 1999]. Results of the mentioned studies show beyond any doubt that
«the missing middle» is actually pretty densely populated by still practiced intermediate exchange
rate regimes [see also Reinhart‐Reinhart, 2003].
«BBC» intermediate regimes reconsidered
«BBC» exchange rates are as close as one gets to relative flexibility of the currency under
evaluation. The main proponent of crawling/monitoring bands and alike regimes has been John
Williamson (1996, 1999, 2000, 2000*). Its name «BBC» owes to the late Rudi Dornbusch, who hu‐
morously summarized regimeʹs main ingredients: a basket (of currencies our currency is pegged
against), a band within which itsʹ floating is limited, and a crawl according to inflation differentials
or other prespecified fundamentals. There are a number of advantages of such a regime. First of
2007 ‐ 20 • Economic Analysis®
all, exchange rate is not tied down but rather able to operate as a proper macroeconomic weather
vane. Nevertheless, a publicly disclosed band could guide market expectations, thereby discourag‐
ing too excessive «noise trading». As opposed to the Krugman‐ian well‐known target zone con‐
cepts, «BBC» exchange rate regimes are much more akin to Jacob Frankel´s crawling band, which
underlines the fact that, need be, the whole corridor should slide as frequent as possible in order to
close the gap between the central parity and updated value of fundamental variables and/or algo‐
rithm chosen to be responsible for equilibrium exchange rate determination. On a top of it, in an
exchange rate regime like this, intramarginal interventions should be prohibited, otherwise poli‐
cymakers would tend to reduce it to de facto crawling peg while the market participants would not
be able to differentiate «noise trading» from persistent «order flows» propped up by ever changing
fundamentals.4 In fact, all of the famous empirical studies which dismissed the mean‐reverting
trait of the exchange rate corridors, for instance [Flood‐Rose‐Mathieson, 1990], [Svensson
1991,1991*] and Neely [1994], dealt with detected «honeymoon effect» although it was a theoretical
violation, whereas majority of existing studies further proposed full price flexibility as an odd
marriage with sticky price reality of contemporary macro. Moreover, Bekaert and Gray (1998) and
slightly later Andersen (2000), by modelling target zones in a discrete time, and by using non‐
parametric techniques and Monte Carlo simulations have clearly showed that exchange rate trajec‐
tory in absence of intramarginal interventions does display mild but nontrivial stabilisation effect
(or S‐shaped recovery in Krugmanʹs own words)! Finally, perhaps the most convincing argument
for mean‐reverting character of crawling/monitoring bands was raised by Williamson (1999), who
verifies that within a currency band (in comparison with free floats) spot rate movements are fol‐
lowed by much less volatile reactions of forward rates.
Inflation targeting in emerging markets
After considerable success which inflation targeting achieved in OECD countries (primarily
in terms of price stability), it is swiftly becoming the holy bible of monetary policy making in de‐
veloping countries and emerging markets, too. Nonetheless, lagging behind the contagious fash‐
ion, there is still barely a handful of serious studies that examine the compatibility of textbook IT
design with transitional peculiarities of emerging markets. In terms of voluminous mainstream
literature, however, supremacy of inflation targeting over thus far identified alternative monetary
strategies is strikingly obvious. Instead of irrelevant or ineffective targeting of monetary aggre‐
gates on one hand, and virtually impossible targeting of nominal output on the other [Eichen‐
green, 2002], inflation targeting concentrates directly on price stability and, if not too rigidly aimed
at, under certain well known preconditions simultaneously stabilises output gap around its natu‐
ral level [Blanchard, 2003]. Moreover, since «we are all dead in the long run», inflation targeting
also has understandable advantage over price level targeting, advocated by Svensson (1997), De‐
hejia and Rowe (2000), and Ball, Mankiw and Reis (2003), since IT regimes tolerate so‐called base‐
drift which enables additional anticyclical flexibility and keeps the focus on price differences
rather than prices themselves.5 On a top of proclaiming mid‐term numerical inflation target for
4 The idea is that honeymoon-free (no marginal interventions), wide enough currency bands leave reasonably sufficient
space and time to identify more robust/durable exchange rate movements and hence distinguish them from temporary
“noise” at FX market. However, it is true that even under such a regime we cannot always be 100% sure regarding the
exact nature of exchange rate dynamics at hand.
5 It is in order to say that, if one ignores the concern about time inconsistency and credibility of central bank’s commit‐
ment to price‐level targeting, superiority of price‐level targeting over inflation targeting could indeed materialize in the
long‐run, yet at the short‐run expense in terms of GDP and inflation volatility, brought about by rectifying the base‐drift
intolerance.
Volume 40 • Spring 2007 • 21
monetary authorities, full‐fledged inflation targeting strategy assumes institutional support and
informational intensity as well as maximum transparency in explaining the concept and operative
policy of the central bank [Mishkin, 2000].
As prerequisites for successful implementation of inflation targeting in transition countries,
we choose to stress the following from the latest (and scarce) literature [Fraga‐Goldfajn‐Minella,
2003], [Jonas‐Mishkin, 2003], [Przystupa, 2002]:
1. Well understood transmition mechanism from exchange rate into inflation as well as
feed‐back influence of instrument rule on inflation.
2. Balanced budget and absence of the so‐called fiscal dominance
3. Reasonably well developed financial system
4. Institutional and actual independence and transparency of central bank
5. Freely floating exchange rate, i.e. absence of any other nominal anchor apart from in‐
flation target.
However, in our opinion, already an academic consensus stated under 5), represents a dog‐
matic misconception and should be ruled out, at least for small open economies in transition with
fragile and emerging markets. Without properly developed hedging instruments coupled with the
lack of domestic savings, speculative attacks on official reserves and, broadly speaking, exogenous
shocks for capital flows cause significant volatility of exchange rate and economic activity if ex‐
change rate is totally neglected by monetary authorities. Moreover, even if exchange rate pass‐
through in emerging markets appears to be dramatically lower than a priori expected,6 many au‐
thors claim that in economies with rich and recent history of monetary instability, exchange rate
stubbornly remains to be the focal point of inflation expectations. At last, but not least, exchange
rate oscillations apparently exhibit strong impact on profitability of firms and sectors of emerging
markets, even when aggregate indicators give reason for self‐confidence [Røisland‐Torvik, 2000],
[Ball, 2001].
As a matter of fact, empirical evidence on exchange rate policies effectively exercised in such
economies shows that monetary authorities knew all too good the dangers of completely giving up
the exchange rate, yet potential or actual malfunctioning of IT regime often came about for com‐
pletely opposite reasons – namely too little floating disguised under politically correct fad: man‐
aged or dirty float [Malovic, 2006].
Be that as it may, a handful of serious studies of inflation targeting in emerging markets de‐
tected several more problems or potential limitations of IT regimes in transition reality [Eichen‐
green, 2002], [Fraga‐Goldfajn‐Minella, 2003], [Ho‐McCauley, 2003], [Jonas‐Mishkin, 2003]: a) credi‐
bility, especially under strict (point) targets and lack of communication and reporting; b) appropri‐
ateness of inflation measure chosen to be targeted; c) too high partial dollarisation of economy; d)
inherent imprecision of inflation forecasting; d) feasibility of softening the edges of inflation band
targeted if the need arises, and specifically e) too rigid –implicit or explicit exchange rate targeting.
Regarding the credibility of inflation targeting strategy, this short experience of emerging
market economies seemingly shows that strict inflation targeting, i.e. taking responsibility for hit‐
ting the point announced might not be very credible after all. Since statistical base and experience
may typically be lacking, whereas administrative control of quite a few prices and ongoing tax
reforms in emerging market countries are evidently still deployed [Jonas‐Mishkin, 2003], it is cer‐
tainly much safer and more credible to commit to targeting inflation within a range than its par‐
6 Indeed, in many transition countries exchange rate pass‐through, initially believed to be close to 1, turned out to be
significantly lower than 1, in some cases even less than 0.5. See [Ho‐McCauley, 2003], [Hausman‐Panizza‐Stein, 1999].
2007 ‐ 22 • Economic Analysis®
ticular level. Even more so should politically motivated conflicts between fiscal expansion and
monetary commitments arise.
In respect to inflation measure one should stick to, we agree with the thin minority of opin‐
ions that –unless any other index has been previously used for sufficiently long time prior to in‐
troduction of inflation targeting ‐ CPI dynamics is the most appropriate measure of inflation that a
bloke in the street really cares about. Having said that, one cannot be unaware of the difficulties to
control some of the components of CPI inflation that seem to be out of direct and immediate reach
of central banksʹ policies, as pointed out by advocates of core inflation concept. However, it is
common knowledge that core inflation concept excludes several important prices with systemic
secondary consequences for the overall inflation. Hence, even if one rules out the possibility that
obscurity is the exact reason for the weapon of choice, core inflation is often at the verge of missing
the point as far as general public is concerned: namely, if you are targeting something that´s not
completely relevant for the behaviour of the market you strive to control, that might easily pinch
some rhyme or reason out of the whole business.
Nonetheless, perhaps principal cause of problems with IT in emerging market countries is
proverbially massive dollarisation of economy and specifically of its external as well as domestic
liabilities. Notorious volatility of international capital flows and «pillow nature» of savings in
emerging markets exacerbate monetary policy challenges further still. According to Fraga, Gold‐
fajn and Minella (2003), typical sequence of events in such a constellation goes like this: «sudden
stop» phenomenon a la Calvo heavily influences nominal (and real) exchange rate, which in turn
via pass‐through effect fuels the inflation, so that consequently restrictive monetary policy cata‐
pults interest rates and thereby inflicts third generation financial crises onto fragile banking cum
corporate sector (both burdened with hard‐currency‐indexed liabilities), instead of bringing about
stabilisation. One way of dealing with the volatility of capital flows is unavoidably a Krugman‐ian
(2000) version of the Tobin tax, i.e. pecuniary check‐point installed on the way in rather than on the
way out. Severe dollarisation problem in many emerging markets, as a flip‐side of the bad old
Greshamʹs law, alas, is likely to consume much more time and credibility effort. However, dollari‐
sation (as a phenomenon, not as an exchange rate regime) is to the extent endogenous to a lagged
observations of the adopted monetary policy regime [Batini et alia, 2006], so that establishing disin‐
flation together with rising export performance through credible policy rules is likely to eventually
crowd out dollarisation to acceptable levels. Probably the key ingredient for success in the afore
mentioned ball‐game rests once again with the ability to reassure market participants that currency
stability and exchange rate fixing are not one and the same thing...
In addition, inherent imprecision of inflation forecasting apparently did not restrain few
emerging market countries from flirting with incomplete yet narrow inflation targeting in the early
days of their IT experience. Nowadays virtually all inflation targeters in emerging markets realise
the benefits of flexible inflation targeting, since targeting a band instead of a point doesnʹt make
central banker less of a sharp shooter, but makes him a wiser one. The latest empirical evidence
implies that hitting the often impossibly strict inflation target may be less crucial for regimeʹs suc‐
cess than the sustained pursuit of improvements and publicly released revision of reasons for vio‐
lating the forecast once the framework has been adopted [Batini et alia, 2006]. Moreover, leaving a
bit more room for targeted inflation enables monetary authorities to acknowledge (if not pursue)
other legitimate goals of macroeconomic policy, economic growth and employment being the first
two coming to economistʹs mind. Goals that, mind you, tend to be even more pressing in emerging
market reality. Therefore, flexible inflation targeting implies a trade‐off between stabilising infla‐
tion around and within the precommitted inflation target and minimising output gap defined as
departure of current output from its natural level or otherwise defined potential.
Volume 40 • Spring 2007 • 23
However, as previously stated, most studies on inflation targeting have tended to ignore the
emerging market reality and open economy issues. Woodford (2003), for instance, in his seminal
book on monetary policy and inflation targeting in particular practically does not deal with open
economy issues and exchange rates at all. In the meanwhile, apart from eight industrial countries
which exercise full‐fledged inflation targeting, there are more than 15 emerging market economies
that opted for either informal or full‐fledged inflation targeting. Majority of those latter ones in
reality opted for –often strict at onset‐ inflation targets coupled with managed floats. Therefore, the
last section examines the feasibility, rationale as well as potential pitfalls of simultaneous exchange
rate and inflation targeting.
«BBC plus» ‐ simultaneous exchange rate and inflation targeting
«BBC plus», catch‐phrase from the title of this section, is actually paroding Goldsteinʹs (2002)
famous «dirty floating plus», which posits the latest professional fad, that managed floating en‐
dures all identified challenges of IT in small open economies and that these two are an item. The
argument goes that since emerging markets countries cannot abstain from exchange rate targeting
even under IT regime, then they should manage exchange rate «under the table», intervene and
manage «silently», «in the dark», without precommitment to any last ditch that would have to be
defended. Thereby we would still be able to avoid speculative attacks, keep central bankʹs reserves
safe while successfully controlling inflation.
But, is there something fishy with the matrimony of dirty float and an inflation targeting re‐
gime? In theory, managed floating plus represents an IT package which should indeed retain all
the desirable features of relatively flexible exchange rate regime while still being able to address
nominal anchor and balance‐sheet issues critical for emerging market economies. However, too
aggressive and fearful interventions in practice fall far from correcting currency mismatches, as a
matter of fact, they often turn into conservation of fragile macroeconomic stability via tying down
interest rates and exchange rate itself, thereby threatening with recessionary consequences the very
(and alas often unrealistically low) inflation target they aim to preserve. In addition to that, espe‐
cially when coupled with too narrow inflation targets,7 dirty floats become, ironically, rather prone
to speculative attacks of the first generation type as explained by Kumhof [2001]. Finally, our prin‐
cipal critique of «managed floating plus» has to do with the efficiency (and effectiveness for that
matter) of sterilised foreign exchange interventions, that are so persistently carried out in emerging
markets on an ever growing scale. Bofinger and Wollmershaeuser (2003), as the most famous ad‐
vocates of managed floating, by dedicating almost a third of their managed floating study to ster‐
ilisation issues, implicitly admit that utility and usefulness of dirty floats depend on the ability to
effectively and efficiently sterilise foreign capital inflows. Moreover, Bofinger and Wollmer‐
shaeuser (2003) recognise that if we care about monetary concept broader than high‐powered
money, as most central banks nowadays do, successful sterilisation of foreign capital inflows is
identically equivalent to interest rate targeting. Now, if we are severely limiting the motion of ref‐
erence interest rate (typically 2W to 3M repo), then we are losing degrees of freedom in trying to
implicitly target exchange rate by targeting interest rate which is by definition a principal instru‐
ment in controlling inflation target!? Let alone the steeply rising quasi‐fiscal costs of any prolonged
sterilisation [Kumhof, 2000], [Kletzer‐Spiegel, 2002]...
7 In essence, strict inflation targets are almost equivalent to hard effective exchange rate pegs, with all the dangers of
committing to such a numerical target in the case of speculative attack on foreign reserves [Calvo, 2000], [Kumhof‐Li‐
Yan, 2001]. If the Ricardian equivalence is violated, while the currency is too heavily “managed”, speculative attack
tends to simply relocate from one nominal anchor and forex market to another (inflation target) and government securi‐
ties market as in [Kumhof, 2001] and [Corsseti‐Mackowiak, 2003].
2007 ‐ 24 • Economic Analysis®
Be that as it may, does that mean that we could abandon exchange rate targeting all to‐
gether in small open economies once the inflation targeting regime is in place? Or is there still a
role for auxiliary anchor, especially having in mind conveniently soft edges of sufficiently wide
monitoring bands designed by Tarapore Committee of the Indian central bank, which allow resort‐
ing to stretching clause in the event of temporary but massive non fundamental shocks [William‐
son, 2000*]? Moreover, until the dirty float come‐back, economists seemed to agree upon the su‐
premacy of well‐designed transparent rules over under the table non‐committed discretion. By not
having the opinion on whereabouts of central parity (nominal and/or real effective exchange rate)
or by hiding that information from the public, do central banks of managed floaters add up to
transparency of IT regime and their interest rules as well as reaction functions? After all, whom are
proponents of dirty floating plus trying to fool, by claiming that ‐in high pass‐through and highly
indebted environment‐ once authorities choose the inflation target and its «maturity», speculators
cannot feel the trend of exchange rate puls even if «Maginot lines» are not formally announced,
since dirty pegged exchange rate is the main worn‐out instrument for controlling reflation? For
ought we care, all the answers to previous rethoric are bound to be negative.
In as much as non‐zero pass‐through from exchange rate affects inflation as well as to the ex‐
tent that any more serious exchange rate misalignment affects output gap, it is painfully obvious
that every emerging market central bank would have to keep a close eye on relationship between
inflation targeting and national currency movements and react through some kind of Taylor rules,
even if there is no publicly announced exchange rate targeting [Clarida, 2001], [Eichengreen, 2002],
[Flamini, 2003]. On a top of it, all prospective candidates for sooner or later E(M)U membership,
including those which have already introduced inflation targeting,8 will have to introduce very
formal and irrevocable exchange rate target zones within the ERM II: target zones for which much
more flexible yet similar «BBC» regimes appear to be a good starting point and an excellent itera‐
tive interim solution.9
Nevertheless, the question of whether exchange rate targeting should be an independent,
explicitly modelled particle of central bank´s reaction function remains a bitterly debated puzzle in
contemporary open economy macroeconomics. Batini et alia (2006) in general, Bufman and Lei‐
derman (1999) and Brenner and Sokoler (2001) in the case of Israel, Jonas and Mishkin (2003) for
Hungary, Morande (2001) and Schmidt‐Hebbel and Tapia (2003) in the case of Chile, to name only
the few, express poorly‐documented concern for possible dynamic inconsistency between (or even
reason for existence of) two nominal anchors. As far as potential conflicts between simultaneous
achievement of inflation target and exchange rate target are concerned, after thorough examination
of Chilean, Israeli and Hungarian case, our conclusion is that eventual conflict between inflation
and exchange rate target can (yet by no means must) happen either when the level and/or width of
target zones is ill‐chosen (often set unrealistically low and narrow), or as a consequence of too
steeply predetermined rate of currency crawl. As a residue, equally serious pitfall for «BBC plus»
and alike policy mix opens up out of controversial ambition of monetary authorities in emerging
markets to overly exploit sterilised intervention in order to simultaneously accomplish several
mutually excluding goals [Malovic, 2006].10 Otherwise, direction if not magnitude of nominal ex‐
8 For example Romania, Turkey and Serbia.
9 It goes without saying that within fixed, non sliding currency band preconceived under ERM II all of the perils of pre‐
committed pegging are once again threatening countries’ exchange rate targets, even more so since the ECB in practice
insists upon much thinner corridors than officially proclaimed, yet one should not forget that responsibility for defend‐
ing such a target zone rests not only with the national central bank in question, but with the ECB itself too.
10 Political leaders often utter the pressing need to attract as much FDI as possible and provide growth while simultane‐
ously urging for decreasing or even overthrowing balance of payments deficits. Similarly, emerging markets frequently
strive to obtain massive capital inflows in prolonged periods yet to avoid either premature debt repayment, their curren‐
cies’ appreciation or increase in inflation as a corollary of abandoning sterilization policy.
Volume 40 • Spring 2007 • 25
change rate movements within the band and inflation dynamics within the targeted corridor ought
to be fairly compatible. Regarding the rationale for their simultaneous existence, one of the best
explanations so far as to why this may be a good idea and in fact may enhance the social welfare
(further minimise central bank’s loss function) is the one given by Edwards (2006, pages 22‐23),
who rightfully realises that when there is a lag in reaction of both inflation and output gap to ex‐
change rate changes, «(...) the central bank may want to pre‐empt their effect by adjusting the pol‐
icy stance when the exchange rate change occurs, rather then when its effects on inflation and out‐
put are manifested».
Therefore, for small open economies with fragile emerging markets central bankʹs loss func‐
tion (that we are trying to minimise) obviously cannot be of the simplest flexible inflation targeting
form:
( )[ ]2121 )(2
1 T
t
TT
t yy −+−=Λ ++ οππ
Eichengreenʹs (2002) and Woodfordʹs (2003) specification which includes a concern for limit‐
ing variability of interest rates observable in many transition countries is perhaps acceptable, but
probably suffers from the same instrumental and operational pitfalls already assigned to managed
floating plus:11
2
1
2
1
2
1 )()(
T
t
TT
t
T
yt iiyy −+−+=Λ +++ εοοπ
Hence, it is advisable for emerging market central bankers to explicitly adopt logarithmic
version of flexible nominal effective exchange rate target following Svensson (2004):12
( ) ( ) +−+−=Λ ++
2
1
2
1
T
t
TT
tt yyoππ (€t+1 ‐ €T)2
Naturally, this kind of objective function must not degenerate into manipulating the real ex‐
change rate. Moreover, even (wide enough) sliding currency corridor let alone monitoring band
would enable “BBC plus” enough flexibility to tolerate transitional nonfundamental exchange rate
fluctuations. At the same time, transparency of exchange rate regime and its stabilisation proper‐
ties would be preserved [Forbes‐Kofman, 1998], without confusion in regard to the primacy of in‐
flation target over auxiliary nominal anchor.
To the extent that nature of exchange rate fluctuations is not readily verifiable, Ball (1998)
suggests another way to avoid the “exchange rate whiplashing”, namely, central bankers could
choose to target longer run inflation, which again depends upon lagged value of currency rate,
although in the instrument rule one gets only the output channel utilized to control inflation.13
However, Ball (2001) implicitly admits that, unlike market participants in Australia, emerg‐
ing markets share the Keynesian view of the long run, by advocating his Monetary Conditions
Index as an optimal instrument rule for small open economies. MCI rule appoints simultaneous
fine‐tuned interest rate and exchange rate management in order to achieve inflation target, with
real exchange rate target on the left side if we assign it primarily the role of instrument (operating
11 For the purpose of sketching several convenient ways of including some other variables (exchange rate being primus
inter pares) into central banks objective function in emerging small open economy version of inflation targeting, we ab‐
stract from legitimate need to refine formulae with precise definition and/or calibration of weights, as well as acknowl‐
edge often autoregressive nature of inflation. For more on those issues consult for example Svensson and Woodford
(2003).
12 Indeed, one could also target real (rather than nominal) exchange rate. Svensson proposes that as an alternative too in
the same presentation.
13 Surely the best yet only theoretical option (out of statistical reach) for most emerging markets would be the ability to
break down output gap into its tradable and nontradable component. Following Svensson (2004), one would mini‐
mize ( ) ( ) ( )2)1(2)1(21 TtNTTtTTtt yyyyo Ν+ΝΤ+ΤΤ+ −+−+−=Λ οππ , with more than obvious implicit presence of
real exchange rate concern once again.
2007 ‐ 26 • Economic Analysis®
target) or on the right side if its limited floating is simply an indicator to be taken into considera‐
tion [Ball, 2001]:
φ iR +(1‐φ )εR = Tϑ ( ) ( )TtTytt ννοππ −+− −1
iR= ( ) ( )Tt
T
y
tt
T
νν
φ
ο
ππ
φ
ϑ
−+− −1 ‐ ⎟⎟
⎠
⎞
⎜⎜
⎝
⎛ −
φ
φ1
εR
In those two mathematically identical formulae, ν represents the output gap while φ indi‐
cates the weight given to real interest rate in the policy mix. Faced with repeatedly occurring Calvo
shocks, simultaneous exchange rate and inflation targeting policy fares better than simple infla‐
tion‐centred interest rate rule: in such a setting, sudden capital flight provokes interest rate hike
which compensates for real exchange rate depreciation caused by the balance of payments disequi‐
librium. In fact, keeping MCI in check lowers domestic spending (via interest rate channel)
whereas increases net export (via real exchange rate channel) and thereby restores balance of pay‐
ments and stabilises output. Nevertheless, it is fair to say that this conventional interpretation of
MCI might reveal inferior stabilisation properties as opposed to Taylor rule in the face of Prebisch‐
adversity (net export revenue shock) [Ball, 2001]. However, more careful alternative specifications,
such as taking care of policy makerʹs preferences and direct exchange rate‐domestic inflation link
as in Guender (2005), as well as rethinking the role of relative prices in effectiveness of exchange
rate as a shock absorber as demonstrated in Edwards (2006), could rule out such deficiencies.
Conclusion
This paper has tried to argue that contrary to Taylor (2001) and officially preached stance of
many central banks in small open economies, exchange rate targeting does play an important role
in the conduct of optimal monetary policy. After acknowledging the supremacy of somewhat for‐
gotten sliding currency bands vis‐à‐vis the still ruling bipolar view, on one hand, as well as vis‐à‐
vis the latest dirty floating fad, on the other, paper turns to specificities of inflation targeting in
emerging markets as well as rationale for versus potential pitfalls of simultaneous exchange rate
and inflation targeting.
Discussion has been closed by reiterating the inevitability of close relationship between infla‐
tion targeting and exchange rate targeting. Hence, several possible reaction functions for the mone‐
tary authorities in emerging markets were suggested, among those already laid out in the related
literature.
References
Andersen, A. (2000), »Was the Honeymoon Effect Effective? An Analysis of the EMS Target Zone«, The Arhuus School of
Business, mimeo.
Ball, L. (1998), «Policy Rules for Open Economies», Research Discussion Paper 9806, Reserve Bank of Australia.
Ball, L. (2001), «Policy Rules and External Shocks», Norges Bank, Arbeidsnotat 3.
Ball, L.‐Mankiw, G.‐Reis, R. (2003), «Monetary Policy for Inattentive Economies», NBER Working Paper 9491.
Batini, N. et alia (2006), «Inflation Targeting and the IMF», International Monetary Fund, Washington, (March).
Bekaert, G.‐ Gray, S. (1998), “Targets zones and exchange rates: an empirical investigation”, Journal of International
Economics Vol. 45, pp.1‐35.
Blanchard, O. (2003), “Comments on ‘Inflation Targeting in Transition Economies‐Experience and Prospects’ by Jiri Jonas
and Frederic Mishkin”, MIT, mimeo.
Bofinger, P.‐ Wollmershäuser, T. (2003), “Managed Floating: Theory, Practise and ERM II”, Deutsche Bank/CEPR Work‐
ing paper, (January).
Volume 40 • Spring 2007 • 27
Brener, M. ‐ Sokoler, M. (2001), “Inflation Targeting and Exchange Rate Regimes: Evidence from Financial Markets”,
Bank of Israel Discussion Paper.
Bufman, G.‐Leiderman, L. (1999), »Inflation Targeting under a Crawling Band Exchange Rate Regime: Lessons from
Israel«, Lecture Notes, chapter 9, mimeo.
Calvo, G. (2000*), “Capital Markets and the Exchange Rate, with Special Reference to the Dollarisation Debate in Latin
America”, University of Maryland at College Park, mimeo.
Calvo, G.‐Reinhart, C. (2000), “When Capital Inflows Come to a Sudden Stop: Consequences and Policy Options,” in P.
Kenen and A. Swoboda (eds), Key Issues in Reform of the Intertional Monetary and Financial System, (Washington, DC: In‐
ternational Monetary Fund).
Calvo, G.‐Reinhart, C. (2000*), “Fear of Floating”, University of Maryland, mimeo.
Clarida, R. (2001), “Empirics of Monetary Policy Rules in Open Economies”, NBER Working Paper 8603.
Corsseti, G.‐Mackowiak, B. (2003), “Fiscal Imbalances and the Dynamics of Currency Crises”, Humboldt Univesity in
Berlin Working Paper, mimeo.
Dehejia, V.‐Rowe, N. (2000), “Macroeconomic Stabilization: Fixed Exchange Rates vs. Inflation vs. Price Level Target‐
ing”, Carleton University at Ottawa, (March), mimeo.
Detken, C.‐Gaspar, V. (2003), “Maintaing Price Stability under Free‐Floating: A Fearless Way out of the Corner?”, ECB
Working Paper 241.
Edwards, S. (2006), “The Relationship between Exchange Rates and Inflation Targeting Revisited”, NBER Working Paper
12163.
Eichengreen, B. (2002), “Can Emerging Markets Float? Should They Inflation Target?”, Banco Central do Brasil Working
Paper 36.
Fischer, S. (2001), »Exchange Rate Regimes: Is the Bipolar View Correct?«, IMF’s Finance&Development, June, Vol. 38,
no. 2, pp. 18‐22.
Flamini, A. (2003), “CPI Inflation Targeting and Exchange Rate Pass‐Through”, Princeton University, (May), mimeo.
Forbes, K.‐Kofman, P. (1998), “Bayesian Target Zones”, mimeo.
Fraga, A.‐Goldfajn, I.‐Minella, A. (2003), “Inflation Targeting in Emerging Market Economies”, Banko Central do Brasil,
Working Paper 76.
Goldstein, M. (2002), “Managed Floating Plus”, Institute for International Economics, Washington DC, passim.
Guender, A. (2005), “On Optimal Monetary Policy Rules and the Construction of MCIs in the Open Economy”, Open
Economies Review 16, Springer Verlag, pp. 189‐207.
Hausmann, R. (2002), “Unrewarded Good Fiscal Behavior: The Role of Debt Structure”, Harvard University, mimeo.
Hausmann, R.‐Panizza, U.‐Stein, E. (1999), “Why do Countries Float the Way they Float?”, Inter‐American Development
Bank, (November), mimeo.
Ho, C.‐McCauley, R. (2003), »Living with Flexible Exchange Rates: Issues and recent experience in inflation targeting
emerging market economies«, BIS Working Paper 130.
Flood, R.‐Rose, A.‐Mathieson, D. (1990), “An Empirical Exploration of Exchange Rate Target Zones”, NBER Working
Paper 3543.
Frankel, J.‐Fajnzylber, E.‐Schmuckler, S.‐Serven, L. (2001), “Verifying Exchange Rate Regimes”, Journal of International
Economics Vol. 66, pp. 351‐386.
Jonas, J.‐ Mishkin, F. (2003), Inflation Targeting in Transition Countries: Experience and Prospects, NBER Working paper
9667.
Kletzer, K.‐ Spiegel, M. (2002), “Sterilization Costs and Exchange Rate Targeting”, UCSC Working Paper, mimeo.
Krugman, P. (2000), “The Return of Depression Economics”, Penguin Books.
Kumhof, M. (2000), “Capital Inflow Sterilization – Through Lower Interest Rates?”, Stanford University, mimeo.
Kumhof, M. (2001), “A Critical View of Inflation Targeting: Crises, Limited Sustainability and Agregate Shocks”, Banco
Central de Chile ‐ Documentos de Trabajo N. 127.
Kumhof, M.‐Li, S.‐Yan, I. (2001), “Balance of Payments Crises under Inflation Targeting”, Stanford University, mimeo.
Malovic, M. (2006), “International Financial Crises”, CID, Faculty of Economics, Belgrade.
Morande, F. (2001), “Exchange Rate Policy in Chile: Recent Experience”, paper prepared for the Conference “Exchange
rate regimes: Hard peg or free floating?”, organized by IMF Institute on March 19‐20, Washington, DC.
2007 ‐ 28 • Economic Analysis®
Mishkin, F. (2000), “Inflation Targeting in Emerging Market Countries”, NBER Working Paper 7618.
Neely, C. (1994), “Realignments of Target Zone Exchange Rate Systems: What do We Know?”, Federal Reserve Bank of
St. Luis, September/October Issue, mimeo.
Przystupa, J. (2002), “The Exchange Rate in the Monetary Transmission Mechanism”, National Bank of Poland,
(downloaded from official internet page of NBP).
Reinhart, C.‐Reinhart, V. (2003), “Twin Fallacies about Exchange Rate Policy in Emerging Markets”, NBER Working
Paper 9670.
Reinhart, C.‐Rogoff, K. (2004), “Modern History of Exchange Rate Arrangements: A Reinterpretation”, Quarterly Journal
of Economics, Vol. CXIX, Issue 1, pp. 1‐48.
Røisland, Ø.‐ Torvik, R. (2000), “Fiscal Policy under Inflation Targeting”, Norges Bank, Arbeidsnotat 15.
Schmidt‐Hebbel, K.‐Tapia, M. (2003), »Toward Floating and Inflation Targeting in Chile«, Central Bank of Chile, (pre‐
pared for CEPT/Deutsche Bank Conference ʹManaged Floating‐ An Alternative to Two‐Corner Solution?ʹ, January 30‐31,
Kronberg, Germany).
Svensson, L. (1991), “The Simplest Test of Target Zone Credibility”, IMF Staff Papers, (September), pp. 655‐65.
Svensson, L. (1991*), “The Foreign Exchange Risk Premium in a Target Zone with Devaluation Risk”, CEPR Discussion
Paper 494.
Svensson, L. (1997), “Price vs. Inflation Targeting: A Free Lunch?”, On‐line paper (taken from Svensson’s Princeton web
page).
Svensson, L. (2004), “Fexible Inflation Targeting: Principles and Possible Improvement”, (presentation given at Norges
Bank, 25. March, Oslo, Norway).
Svensson, L.‐Woodford, M. (2003), “Implementing Optimal Policy through Inflation‐Forecast Targeting”, (paper pre‐
sented at NBER Conference on Inflation Targeting, 23‐25 January, Bal Harbour, Florida), mimeo.
Taylor, J. (2001), “The role of the Exchange Rate in the Monetary Policy Rules”, The American Economic Review 91, pp.
263‐267.
Williamson, J. (1996), “The crawling band as an exchange rate regime: Lessons from Chile, Colombia and Israel”, Insti‐
tute for International Economics, Washington, DC.
Williamson, J. (1999), »Crawling Bands or Monitoring Bands: How to Manage Exchange Rates in a World of Capital
Mobility«, Policy Briefs, No. 99‐3, Institute for International Economics, Washington DC.
Williamson, J. (2000), “Designing a Middle Way between Fixed and Flexible Exchange Rates”, (paper presented on con‐
ference “Monetary and exchange rate policy: options for Egypt” that took place in Egyptian Centre for Economic Stud‐
ies, 19‐20 November, Cairo).
Williamson, J. (2000*), “Exchange Rate Regimes for Emerging Markets”, Institute for International Economics, Washing‐
ton, DC.
Woodford, M. (2003), “Interest&Prices‐Foundations of a Theory of Monetary Policy”, Princeton University Press.