Brazilian Journal of Political Economy, vol. 19, nº 3 (75), pp.

426-441, July-September/1999

Was there an alternative to the Brazilian crisis?

Houve uma alternativa para a crise brasileira?

J. A. KREGEL*

RESUMO: Diferentemente da Ásia, altas taxas de juros e taxas de câmbio estáveis associa-
das ao Plano Real não produziram deflação da dívida corporativa devido ao baixo endivi-
damento corporativo no Brasil. Em vez disso, altas taxas de juros fizeram com que os saldos
externo e fiscal se deteriorassem, reduzindo a confiança. Qualquer tentativa de reduzir as
taxas de juros trazia a ameaça de fraqueza da moeda e o risco de inflação. A crise deveu-se
à dependência de altas taxas de juros para atrair fluxos de capital insuficientes para produ-
zir investimentos que proporcionavam uma taxa de crescimento satisfatória.
PALAVRAS-CHAVE: Crises financeiras; fluxos de capitais; Plano Real; globalização.

ABSTRACT: In difference from Asia, high interest rates and stable exchange rates associated
with the Real Plan did not produce a corporate debt deflation because of the low corporate
indebtedness in Brazil. Instead high interest rates caused both the foreign and fiscal balances
to deteriorate, reducing confidence. Any attempt to reduce interest rates brought the threat
of currency weakness and the risk of inflation. The crisis was due to the reliance on high
interest rates to attract capital flows which were insufficient to produce investment which
gave a satisfactory rate of growth.
KEYWORDS: Financial crises; capital flows; Plano Real; globalization.
JEL Classification: F41; F65.

I. INTRODUCTION

There are two basic approaches that can be taken to providing alternative
views about how the global trade, production and financial system can be made
more stable. One is based on the presumption, frequently advanced by the US
Treasury, that the global allocation of capital is both efficient and stable when
capital is free to seek the highest available returns without impediment or admin-
istrative restriction. If the system is unstable, it is because countries are insuffi-

* Professor in the Department of Economics of the Università degli Studi di Bologna and Adjunct Pro-
fessor of International Economics in the Johns Hopkins University Paul Nitze School of Advanced In-
ternational Studies. E-mail: [email protected].

426 • Revista de Economia Política 19 (3), 1999 http://dx.doi.org/10.1590/0101-31571999-1189

ciently prepared, their economic institutions are not sufficiently robust, or because
their domestic policy choices are incoherent and contradictory. The solution that
is proposed is to strengthen the economic institutions and to ensure sustainable and
coherent macroeconomic policies.1
The second approach is that there are systemic, endogenously generated insta-
bilities in an economic system based on free goods and capital markets which can-
not be remedied by means of micro solutions, such as writing better debt contracts
or providing better regulation and supervision of financial institutions. This is the
position adopted in this paper. It argues that the financial crises that have taken
place since the late 1970s are systemic, and as such can be expected to continue.
Making the system more stable means changing its basis in free, unhindered flows
of capital across countries.

II. THE LESSONS OF STABILIZATION IN THE 1990’S

One of the most remarkable aspects of the recent Asian crisis was the good
macroeconomic fundamentals of those countries involved in the crisis as compared
with the Latin American countries that had previously been the victim of repeated
financial crises. This led analysts to concentrate on the differences in the character-
istics of the Asian crises and the Latin American crises and to criticism of IMF
policies as inappropriate to countries facing rapid reversals of capital flows. It has
become common to characterize this difference as one between “current” account
and “capital” account crises. The ideals that countries with weak macroeconomic
fundamentals, viz. fiscal laxity backed by central bank monetization of government
debt, leads to high inflation and excessive current account deficits that eventually
erode the credibility of the exchange rate and produce speculative attacks leading
to an exchange rate crisis. Countries with fiscal rectitude and independent central
banks will have low inflation experience and this good performance attracts large
capital inflows that eventually produce current account deficits that erode the cred-
ibility of the exchange rate and leads to a reversal of capital flows that produces
an exchange rate crisis.2

1 The case for capital account liberalization is a case for capital seeking the highest productivity
investments. We have seen in recent months in Asia – as at many points in the past in other countries
– the danger of opening up the capital account when incentives are distorted and domestic regulation
and supervision is inadequate. Inflows in search of fairly valued economic opportunities are one thing.
Inflows in search of government guarantees or undertaken in the belief that they are immune from the
standard risks are quite another. The right response to these experiences is much less to slow the pace
of capital account liberalization than to accelerate the pace of creating an environment in which capital
will flow to its highest return use. And one of the best ways to accelerate the process of developing such
a system is to open up to foreign financial service providers, and all the competition, capital and
expertise which they bring with them. The recently concluded global financial services agreement
demonstrates that countries recognize these beneficial effects of external liberalization”(Summers, 1998).
2 lt is now common to identify this difference as between current account and capital account crises. I

Revista de Economia Política 19 (3), 1999 • pp. 426-441 427

 Certainly, there were substantial differences in the economic fundamentals of
the Asian countries that experienced crisis in 1997 and the conditions of the coun-
tries that experienced crisis in Latin America in the 1980s and 1990s, yet there were
some remarkable similarities in their domestic policies and the behavior of the
exchange rate.• 3 Further, it is unclear whether the IMF policies were any more ap-
propriate to Latin American current account crises caused by inappropriate domes-
tic policy than they were to the case of the Asian crises that were based on exces-
sively volatile capital flows. Indeed, the Latin American economies that have been
victim of financial crises in the 1990s have been those who have been most active
in their adoption of IMF style stabilization policies. Against this background I want
to argue that it is the domestic stabilization policies, whether introduced autono-
mously or under pressure of the need to borrow from the IMF, that are inherently
destabilizing in a world of globally mobile financial or productive capital invest-
ments. It is for this reason that it is imperative to find an alternative approach to
domestic stabilization policy for the emerging economies if free capital flows are
to become part of the new global order.
The dependence on capital flows as an integral part of such stabilization pro-
grammes means that monetary policy is called upon to play a role for which it is
not designed. The result is excessively high real interest rates that change the struc-
ture of government expenditures away from current account expenditures toward
capital account expenditures, and nullify the operation of some standard econom-
ic stabilizers. These come in two forms, one is the operation of the standard Keynes-
ian multiplier and the other is the role of monetary policy itself. First, when interest
payments become a substantial proportion of government expenditures4 (and com-
pose the entire government deficit) the operation of the so-called Keynesian auto-
matic stabilizers is disrupted (see Kregel, 1994-95). This is aggravated if there is an

draw a rather different distinction in “East Asia is not Mexico: The Difference between Balance of
Payments Crises and Debt Deflation”, in Tigers in Trouble, Jomo, K.S. (ed.), London, Zed Books, 1998;
the difference between a capital account crisis and a debt deflation process is drawn in “Yes, ‘lt ‘ Did
Happen Again – A Minsky Crisis Happened in Asia,” Jerome Levy Economics Institute Working Paper
# 234.
3 There is one aspect of the crisis in Asia reminiscent of the experiences in Latin America in the 1980s
and 1990s. That is the rapidity with which the Thai devaluation spread to other seemingly fundamentally
strong economies in the region. Just as economists insisted on the fundamental strength of Indonesia as
Thailand collapsed, those with a long memory will remember similar statements made about Brazil as
Mexico defaulted in 1982. Just as political leaders and economic policy makers in Asia were slow to
accept the possibility of contagion from the Thai crisis because they believed (perhaps justifiably on the
basis of IMF article 4 consultations) that their economies were fundamentally strong, the same reluctance
was present in Brazil’s reaction to the 1980s crisis. Indeed, it is interesting to note that just as Indonesia
contributed to the Thai rescue in July of 1997, Brazil was one of the creditors represented in the London
Club when Poland declared it was unable to meet its debt service commitments in the Spring of 1982.
It is also mundane to the argument made below that Korean banks were large, leveraged holders of
Brazilian bonds at the outbreak of the Asian crisis.
4 Brazil’s
Federal government is currently spending more on interest payments on the public debt than
on government wages and salaries.

428 Brazilian Journal of Political Economy 19 (3), 1999 • pp. 426-441

implicit money supply rule since this means that interest has to be automatically
rolled over into increasing amounts of debt, virtually eliminating the impact of
government expenditure on aggregate demand (cf. UNCTAD, 1995, part II, chap-
ter IV). This reinforces the corollary proposition that the ratio of debt to income
cannot be stabilized when the real rate of interest on the debt is higher than the
real rate of growth of income since it moves these two variables in the opposite
(and destabilizing) direction. Finally, there is the point that was first raised by Sar-
gent and Wallace that in such conditions, it becomes impossible to impose a mon-
ey supply rule since the payment of interest will be causing deposits to grow at a
rate that exceeds the desired money target growth rate. All of these propositions
are applicable to stabilization policy in any developed country. Yet, they seem to
have avoided them, while developing countries have not.

III. THE ANATOMY OF THE ENDOGENOUS

INSTABILITY OF STABILIZATION POLICIES

The main similarity of recent financial crises is that they appear to have been
initiated by a sharp reversal in capital inflows that seems to be generated by an
endogenous process of deterioration of economic conditions caused by the capital
that has flowed into the country in response to successful application of internal
stabilization policies. This produces sharp divergences in either interest rates or
growth rates relative to the rest of the world. In a word, successful stabilization
policy carries the seeds of its own destruction in the form of excessive capital in-
flows to take advantage of the return differentials. This is a rather different ap-
proach than that which is now commonly adopted by mainstream economists
critical of the traditional explanation of the crises, i.e. self fulfilling expectations. It
also differs from the explanation that suggests that the liberalization of markets
that is part of most stabilization programmes raises the expected returns to invest-
ment in the country because of the anticipated improvement in economic effi-
ciency and thus on the return to capital invested in the country. Rather, the in-
creased flows appear to be more closely linked to the removal of controls on
capital account and the introduction of restrictive monetary policy, privatization,
movements in foreign interest rates and diversification by international institu-
tional investors. In the 1980s, when banks earnings were suffering from domestic
competition, they looked to international investments to increase return on equity
and market share. Again, as yields in developed countries fell in the 1990s, compe-
tition between institutional money managers increased investments in emerging
markets in order to offer their clients attractive yields. In both cases, it was more a
fall in domestic returns in conditions of increasing domestic competition than the

Revista de Economia Política 19 (3), 1999 • pp. 426-441 429

increased returns in developing countries that has provided the motor behind the
capital flows that have been attracted by these policy changes.5
There are a number of factors which could be outlined, but the most important
is the introduction of tight monetary policies and increased freedom for foreign
investors that created interest rate differentials that were exploited by investors.
The existence of large arbitrage flows taking advantage of large international inter-
est rate differentials has been an integral part of each experience of rapidly rising
capital inflows that has preceded financial crises.
It is a characteristic of arbitrage that it should not be permanent; that it is
eliminated in its pursuit. Yet the interest rate differentials that attract foreign capi-
tal inflows are interest rates of corporate debt determined by the market-determined
rates, they are usually the result of economic policy decisions that influence rate
structures. It is now common to say that the flow of capital in free international
markets punishes unsustainable policies. However, the interest rate divergences that
produce these flows in the case of financial crises have usually been associated
monetary policies designed to eliminate domestic inflation, or are linked to “struc-
tural adjustment” policies that are designed to prepare a country for participation
in the international trading system by reducing inflation rates and balancing gov-
ernment budgets.
Unlike traditional arbitrage, there is little reason to believe that capital flows
based on the arbitrage of policy created interest rate differentials will be self-elim-
inating; indeed they may be self-reinforcing.6 When this is the case such arbitrage
flows make the pursuit of domestic stabilization more difficult. For example, inter-
est rate differentials caused by the introduction of tight fiscal and monetary policies
designed to combat domestic inflation that attract foreign arbitrage flows increase
domestic credit and liquidity and thus offset the restrictive monetary policy unless
they are sterilized. If the inflows go to the private sector, sterilization requires issu-
ing more domestic debt, if the foreign inflow is a direct purchase of government
debt it represents borrowing foreign exchange reserves. In either case, it means
borrowing at the domestic interest rate to acquire foreign funds that can only be
invested at the lower foreign rate. This is “negative carry”, i.e. a loss, for the central

5 I have elaborated this theme in “Flujos de capital, Banco Mundial y crisis financiera depués de Bretton

Woods,” Comercio Exterior, vol. 49, nº. 1, enero 1999, pp. 7-15.
6 According the basic foreign exchange market theory, interest rate differentials should be offset by risk
adjustments comprised mainly of expected changes in future exchange rates. According to the interest
rate parity theorem, interest differentials should be offset by forward discounts or premia – or expected
depreciation or appreciation – that offset the differential. Persistent capital flows suggest that this
process is not operative. Many have suggested that it is the pegging of the exchange rate that makes this
so, and that the negative carry is the opposite side of the profit earned on the interest arbitrage. However,
it occurs even in the absence of formal pegging of the exchange rate. This way of looking at things
highlights the inconsistency of stabilization policies and free capital inflows, for the high interest rate
differentials that are required to keep domestic monetary conditions tight produce capital inflows that
create stability in the exchange rate that precludes elimination of the arbitrage profit and any expectation
of the depreciation of the currency large enough to offset the interest arbitrage.

430 Brazilian Journal of Political Economy 19 (3), 1999 • pp. 426-441

bank or the Treasury, and it is important to note that it represents the other side of
the positive carry that comprises the arbitrage profits for the foreign investors.7
Although this is investment at a loss, the inicial impact of such a policy will be
a form of window dressing in which domestic reserves increase and reinforce the
impression that the currency is strong which is the result of the capital inflows.
However, the losses from the negative carry will eventually emerge as an increase
in the budget deficit, leading to an increase in outstanding public debt. If there are
direct sales of debt to foreign holders, the interest payments represent an accumu-
lating claim by foreigners that will emerge in a deteríoration of the services ac-
counts of the balance of payments. If the process of sterilization puts additional
pressure on interest rates the interest rate differential is increased rather than re-
duced and leads to more arbitrage inflows.
Unless the inflation rate is brought down very rapidly the upward pressure on
the exchange rate will mean a real appreciation that damages domestic competitive-
ness and causes a deterioration in the trade balance that will reinforce the deterio-
ration in the services account caused by direct holdings by foreigners of domestic
debt. The strength of the exchange rate also makes restructuring of domestic indus-
try towards exports more difficult just as high interest rates make it more costly
and low internal demand growth makes internal funds less abundant, countering
the required shift in domestic production towards tradables and higher technology
internationally oriented production. This also makes foreign domestic investors a
more necessary part of the restructuring of the economy since they do not face
domestic capital costs nor rely only on domestic markets and have no vested pro-
duction facilities.
The anticipation of successful stabilization of the domestic economy created
by the introduction of an active restrictive domestic policy in the presence of in-
creased openness of goods and capital markets thus produces capital inflows that
lead to a deterioration in the current account and fiscal position and an apprecia-
tion of the real exchange rate. Despite the increasing reserve position of the central
bank, this is clearly an unsustainable policy combination. A change in interna-
tional interest rate differentials, an external shock, some weakness in the exchange
rate, or a butterfly flapping its wings in the Galapagos, may cause changes in for-
eign investors’ assessment of the success of the stabilization policy that leads to a
decline and then a sharp reversal of capital flows. As long as individual economies
require different adjustment policies because of different internal conditions or
because of asymmetry in the international business cycle, global capital flows will

7 In the simplest case, ignoring spreads, the foreign arbitrager borrows at the dollar interest rate and
buys government securities making a profit on the difference. The Central Bank can invest its increased
dollar reserves at the dollar rate, taking a loss on the difference. In the case of sterilization the foreign
arbitrager buys foreign currency and uses it to buy assets from domestic residents. The Central Bank
then has to recover the domestic currency by selling bonds to domestic residents. The arbitrager’s profit
is then the difference between the dollar rate and the domestic asset return, while the Central Bank’s
loss is the difference between the domestic bond rate and the dollar Treasury bill rate.

Revista de Economia Política 19 (3), 1999 • pp. 426-441 431

arbitrage across the interest rate differentials created by these policy differences.
The arbitrage will eventually render the policies unsustainable, rather than eliminat-
ing the arbitrage because the success of the policies allows interest rates to be re-
duced and the interest differentials to be eliminated. It is often stated that the
outflow of funds causes the unsustainability of domestic policies to be exposed. But,
it is important to recognize that this unsustainability has often been induced in large
part by the impact of the flows themselves on the operation of the domestic stabi-
lization policies.
One component that usually accompanies the reversal of flows is the discovery
of structural weakness in the governance and regulation of the banking system.
Large capital inflows usually create over extension in bank lending that creates
difficulties when the flows are rapidly reversed; the result is instability or a collapse
of the banking system. This instability in the banking system is usually blamed on
inappropriate domestic regulation of the financial sector or lax supervision of the
implementation of regulations. While it is certainly true that regulators and super-
visors are notoriously slow to adjust to changes in the structure and activity in fi-
nancial markets, there is no known case in any country, developed or developing,
of a large increase in flows into the banking sector not leading to an over extension
of lending, a decline in the quality of assets and increased laxity in risk assessment.
This was as true of the Florida banks in the 1927 property lending crisis as it was
of US banks and thrift institutions in the late 1980s.
Lending excesses are in general caused by sudden large increases in inflows of
funds or from recent liberalization of controls on acceptable areas of lending activ-
ity. Thus in a number of crises the increase in bank lending was the result of finan-
cial liberalization that allowed banks to enter areas of activity that had been forbid-
den; the increased lending serving primarily to finance a rapid increase in asset
prices and producing a so-called asset bubble that made returns appear higher than
in other areas of the globe and further increased the attractiveness of the country
to foreign money managers. But over-lending, extension of lending to property and
finance companies and the associated bubbles in asset prices was not the only fac-
tor leading to collapse of banking sectors in most crises. An equally important
factor was the existence of substantial currency mismatches on corporate and fi-
nancial balance sheets. This is also linked to the existence of interest rate differen-
tials and is just another aspect of the destabilizing nature of large capital inflows.
In most crises the capital inflows were not only directed to government debt,
but also served to fund domestic lending, either intermediated through the banking
system or directly by domestic firms borrowing in foreign currency terms. It was
motivated by the size and persistence of existing interest rate differentials. It is
important to note that this mismatching not only existed on corporate balance
sheets, but also in the household sector, as the liberalization of the financial sector
and the goods sector led to financial institutions offering consumer credits that had
not previously available to households.
The existence of foreign currency exposure represents two risks to stability.
One is interest rate risk, for the duration of the foreign lending is usually of short

432 Brazilian Journal of Political Economy 19 (3), 1999 • pp. 426-441

duration and will be governed by international interest rates. It may then require
that corporates or households are able to make rapid repayment and for there to
be substitution with domestic credit, which conflicts with the restrictive stabiliza-
tion policy.
The second is exchange rate risk. Large changes in exchange rates can produce
rapid changes in the domestic currency value of liabilities, without any offsetting
change taking place in asset values for banks or companies or households without
sources of foreign earnings. A depreciation of the exchange rate thus leads to an
instant capital loss for banks, firms and households and a concomitant decrease in
capital and wealth. A sufficiently large exchange rate change may not only produce
illiquidity – the inability to realize the value of assets to meet liabilities, but also
insolvency – the shortfall of the value of assets to cover liabilities at existing realiza-
tion prices. Since a rapid capital outflow will usually lead to a depreciation of the
exchange rate, there will be an automatic increase in the fragility of banks’, firms’
and households’ balance sheets.
The only way to stem these losses will be to repay the debt, or to hedge the
remaining risk. This requires purchase of foreign currency and places further pres-
sure on the exchange rate. Thus in those cases where there is substantial foreign
currency lending by banks and borrowing by firms and households, capital out-
flows are usually accompanied by sharp movements in exchange rates that produce
bank failures and corporate bankruptcies.
This brings up the question of the exchange rate regime, and the adjustment
of that regime between stable and floating rates. Indeed, a number of economists
place their interpretation of financial crises on the inappropriate maintenance of
fixed exchange rates. However, for most countries the experience of stabilization
has been that the stability of the exchange rate is often the result of nothing more
than the strength of the flows themselves and the build up in the size of the bor-
rowed reserves that is created through the sterilization that is required to preserve
domestic stabilization policy. They may also be conditioned by the position of the
dollar; for example in the Latin American crisis the dollar was chronically weak
throughout the inflow period. This creates another self-reinforcing process in which
the inflows create the exchange stability that makes foreign lending appear less
risky, as does the build up of foreign currency lending by banks and borrowing by
firms. It is this second aspect that is important in converting capital outflows into
a full scale crisis since a collapse of the exchange rate regime, ora sharp change in
the value of a foreign currency as in the case of the appreciation of the US dollar
in the early 1980s, not only produces a withdrawal of capital, it has a direct effect
on the financial stability of banks and firms and leads to additional speculative
pressure on the currency. Thus, the decision to sell the local currency is driven by
the impact of devaluation on the balance sheets of economic agents, rather than on
their expenditure decisions. With mismatched balance sheets additional losses can
only be avoided by either hedging or repaying, both of which means selling the
currency in the amount of the original borrowing. This will obviously produce
unbalanced market conditions and produce exchange rate adjustments that are far

Revista de Economia Política 19 (3), 1999 • pp. 426-441 433

in excess of what might have been considered sufficient to produce an adjustment
in the foreign balance and speculation of further falls dominate the market.
Thus, the similarities of the recent experience of financial crises appear to
initiate from the introduction of a domestic stabilization policy in which money
growth is restricted, fiscal budget deficits are cut, and privatization of state enter-
prises is introduced along with reduction of barriers to foreign competition in
goods markets, liberalization of capital flows and deregulation of the financial
system along with the use some sort of foreign exchange rate anchor via an adjust-
able peg exchange rate. The tight monetary policy creates an interest rate differen-
tial sufficiently large to attract arbitrage capital inflows that turns the adjustable
peg currency regime into a quasi-fixed exchange rate. The deregulated banks are
free to lend internally in areas such as real estate and consumer finance that had
previously been closed to them, and banks and domestic firms are free to borrow
abroad, thereby reducing the capital costs created by the high domestic interest
rates, but some of this advantage is offset by a usually underestimated exchange
rate risk exposure. The ensuing real appreciation of the exchange rate weakens the
foreign balance, while the attempt to sterilize the inflows leads to even higher inter-
est rates and weaker internal conditions, as well as higher deficits as high interest
rate domestic bonds are issued to sterilize the inflows that are invested at low
foreign rates. Increasing foreign ownership of domestic debt increases the deficit
on services account and eventually either the foreign balance or the fiscal balance
no longer responds to the government policy, or the domestic conditions deteriorate
substantially as demand is compressed, creating conditions in which a rise in for-
eign rates or a fall in domestic rates leads to an outflow of capital, a collapse of the
exchange rate and a massive capital loss on the balance sheets of banks and firms
carrying unhedged foreign currency exposure. In an attempt to cover these losses
the demand for foreign exchange creates massive imbalance and a free fall in the
currency, producing widespread corporate bankruptcy and bank failures. The crisis
is not due to unsustainable policies, but with an attempt to introduce sustainable
policies in conditions of completely free capital markets. It ends with the impossibil-
ity of the success of those policies due to the impact of capital flows. Such crises
have occurred when stabilization has been introduced as a response to domestic
instability via an IMF structural adjustment programme, or as an endogenous re-
sponse to conditions of hyperinflation, as in Brazil, but they have also occurred
when such policies have been practiced with success over substantial periods, as in
Asia. The problem is not with the policies involved, or whether the crisis is a current
account or a capital account crisis, the problem is systemic, and involves something
like a winner’s curse. A country that manages to introduce a stabilization policy
that looks as if it will be successful will win the prize of large capital inflows which
make the continued pursuit of the stabilization policy destabilizing and eventually
produce a reversal of the flows and a financial, banking and exchange rate crisis.

434 Brazilian Journal of Political Economy 19 (3), 1999 • pp. 426-441

IV. WAS BRAZIL’S REAL PLAN FOR STABILIZATION
ENDOGENOUSLY UNSTABLE?

To what extent does this explanation apply to Brazil? And why has Brazil
managed to avoid a full-scale financial crisis? Brazil offers a good example of the
difficulties caused by an excessive reliance on interest rates in the introduction of
stabilization policies. Indeed, Brazil attempted a series of such policies, starting with
the Cruzado Plan through to the Real Plan, to bring inflation under control. Unlike
many developed countries, Brazil did not take these steps because inflation was
impeding growth, indeed the real growth rate was over 7% in 1986 when these
plans were first started, and has never returned to that level since. Further, there
was little problem with the foreign balance. After the 1980s crisis the trade balance
remained in surplus from 1983 until after the introduction of the Real Plan in 1994.
This was primarily due to the use of a flexible exchange rate policy designed to
preserve export competitiveness. Although Brazil had a large public sector debt of
around 50% of GDP in 1986, it had not been created by a buildup of a large stock
of private foreign assets (in many countries this was primarily through capital
flight), but rather was used in large part for the funding of the 2nd National De-
velopment Plan which started in the 1970s to strengthen the internal productive
structure of the economy and provided increased export capacity. Like most other
heavily indebted Latin American countries in the 1980s, the size of the public debt
was increased through the nationalization of virtually all external debt after 1982.
However, the evolution of Brazilian internal debt was strongly influenced by the
persistence of hyperinflation. Falling fiscal revenues due to lags between assessment
and payment were in part offset by a financial transactions tax, while the lags be-
tween budget allocations and expenditures ensured endogenous reductions in real
spending. The result was a falling ratio of net public sector debt to GDP to around
30% by the time of the Real Plan.
With a rapidly declining real value of debt, it is difficult to convince the private
sector to hold it, and the solution was found in the eventual introduction of fully
indexed bills with high liquidity. The logical consequence was that as long as the
government required financing, there could be no money supply policy independent
of this objective. Indeed, given the full indexing, there was a preference for govern-
ment debt over currency, and much as occurred in Italy in the same period, the
former might be said to have formed the real money supply. As a result fiscal pol-
icy was not a viable policy tool, nor was money supply control, leaving the interest
rate as the only policy tool. This had a number of consequences that are not in the
annals of conventional economics, but are well known to economists from coun-
tries that have experienced high and persistent or inertial inflation. First, interest
rates themselves become a major component of governmem expenditure and thus
of the government deficit. This means they become both a direct and an indirect
source of inflation. Further in Latin America in the 1980s the level of interest rates
relative to the rest of the world had virtually no impact on capital inflows, since
capital markets were characterized by what is euphemistically called reverse net

Revista de Economia Política 19 (3), 1999 • pp. 426-441 435

resource transfers – i.e. net capital flows were negative so that Brazil was providing
resources to the rest of the world and in particular to the US. There was also no
impact of high interest rates on the exchange rate, given the size of the inflation
differential and the policy of preserving competitivity through devaluation.
The important point of difference from other Latin American economies is that
Brazil was not building up an excessively large imbalance on foreign account, nor
was the exchange rate becoming overvalued, net public debt was a share of GDP
was declining, and while low from historical perspective, its growth rate was on
average above other economies in the region. However, to achieve this result in
conditions of hyperinflation and full indexing meant abandoning active fiscal pol-
icy and this implied the impossibility of controlling monetary aggregates. The only
policy tool available to try to stem the hyperinflation was interest rate policy. How-
ever, this only reinforced the hyperinflation by causing:
• a direct increase in the costs of capital since there was no long-term capital
market;
• an increase the government deficit, since the outstanding debt was directly
linked to short-term rates;
• an increase in the rate of inflation through the impact on capital costs and
on the fiscal imbalance.
It would be an inappropriate simplification to say that all of Brazil’s difficulties
in breaking out of its hyperinflation were caused by the excessive dependence on
the interest rate as a policy tool. However, the reliance on high interest rates not
only reinforced the inflation that spiraled into indexed inertial hyperinflation, but
as in many other similar cases, it impeded the full development of private long-term
capital markets found in developed countries. Thus when liberalization called the
system of government directed development financing into question, there was no
structure available to take its place. The financial system lived the life of a rentier
on the float created by the adjustment lags in the indexing system of financial
contracts. Indeed, there was hardly any long-term business financing to be done.
Only the State continued to invest in any appreciable magnitude and this pecu-
liarly Brazilian characteristic of efficient State financing of investment was under
increasing attack from the rapid deterioration of government finances and the push
towards increased liberalization from the multilateral financial institutions and the
Collor government.
The key to breaking out of what was becoming hyperstagflation was thus to
reduce the reliance on interest rates as a policy tool and to reduce the level of inter-
est rates. Two factors would initiate this process. One was the liberalization of
foreign trade introduced the Collor regime in 1990, the other was the return of
capital flows to Latin America at about the same time. Led by Mexico’s new eco-
nomic policy introduced by the Salinas government in 1988, capital flows returned
to Latin America after the introduction of the Brady solution to indebtedness. This
solution was basically built on the idea that Latin America would never be able to
repay its debts to commercial banks, so that it would have to borrow in private
international capital markets to do so. In Brazil international foreign exchange

436 Brazilian Journal of Political Economy 19 (3), 1999 • pp. 426-441

reserves more than doubled between 1991 and 1992 despite the fact that the Col-
lor Plan had not succeeded in reducing the hyperinflation to even two digit levels.
These inflows formed the counterpart to the deterioration in the trade account that
was produced by the increased in imports and reduced exports as foreign competi-
tion displaced local producers as the process of tariff reduction and liberalization
of foreign trade was accelerated.8 Paradoxically, it was this process of liberalization
that provided the background that allowed for the success of the Real Plan.
The Real Plan was in many respects similar to the earlier reform plans, includ-
ing the elimination of indexing of wages and prices as a major component, with the
nominal exchange rate the anchor for price stability. And interest rates continued
to be the major instrument of economic policy. What had changed was that the Real
Plan was introduced in the new post-Brady world of restored capital flows and the
developed country obsession with portfolio diversification and excess returns in
emerging markets. In this new context of increasingly free global capital flows inter-
est rates now had an impact on both capital inflows and on the exchange rate.
Indeed, the early period of the Real Plan saw an appreciation of as much of
15% in the Real-dollar rate, which was intended to have been maintained at dollar
parity, and contributed to the fall in prices. The introduction and early success of
the Real Plan thus reinforced the return of net capital inflows, and caused a dra-
matic reversal of the exchange rate policy of the previous period; the real appre-
ciation of the Real produced a growing payments imbalance in the new context of
liberalized foreign trade.
The foreign capital flows that matched the growing foreign imbalance also had
an impact on fiscal conditions, since the Central Bank adopted a policy of steriliza-
tion of inflows in order to protect its inflation fighting monetary policy, thereby
creating “negative carry” on its foreign exchange reserves that fed directly into the
current budget deficit. This deterioration in the fiscal positions was reinforced by
the impact of the rapid decline in the inflation rate. Just as banks had been able to
make politically influenced loans in the knowledge that inflation would quickly
shrink them from the balance sheet, government had been able to make expenditure
commitments that were rapidly reduced in real terms. The absence of inflation
meant that the rate of growth of fiscal revenues fell while programmed indexed
expenditures were slower to adjust, producing an increase in real expenditures,
some of which was visible in increased household incomes, producing a fall in the
saving rate and financing a consumption boom.
The rapid fall in inflation thus created rising real incomes as prices slowed
more rapidly than other asset prices and the prices of more liberalized imports fell
with exchange rate appreciation. This led to a rapid increase in imports. Finally
banks were forced to find real lending business to replace inflation-float, and start-

8 DelfimNetto (1998) notes that Brazil went from being a net exporter to being a net importer of cotton,
with Arge tina being the main beneficiary.

Revista de Economia Política 19 (3), 1999 • pp. 426-441 437

ed to lend aggressively to consumers in the absence of any quick pick up in business
borrowing, adding further fuel to the consumption boom driving imports.
Thus, while the Real Plan was immensely successful in eliminating inflation, it
had not solved the basic imbalance of the economy represented by the reliance on
high interest rates as the single policy tool. Just as high interest rates had reinforced
the hyperinflationary tendencies in a fully indexed economy, high interest rates re-
inforced the imbalances on foreign and domestic account and produced an endog-
enous movement towards deflation in the economy.
First, it is important to note that the success of the Real Plan was buttressed
on the exchange rate stability that was produced by the steady inflow of foreign
capital; the plan thus virtually instituted a permanent policy of high relative inter-
est rates and a large interest rate differential with the rest of the world since any
attempt to reduce rates ran the risk of increasing vulnerability to foreign shocks.
Given the size and structure of government debt and expenditures, this meant per-
manent fiscal imbalance. The high interest rates that were required to ensure the
capital inflows that kept the exchange rate stable, along with trade liberalization,
produced a rising import imbalance that was countered by sterilization of the in-
flows which increased the negative carry on foreign reserves and thus increased the
government deficit and thus outstanding debt. Any attempt to offset this vicious
circle would require lower government expenditures, which along with the high
interest rates would led to depressed internai demand and lower growth – as well
as lower tax yields. The Real Plan could not restore Brazilian growth with price
stability because it contributed a distorted internal public debt structure and could
not have adopted an expansionary monetary policy even if the government had
desired one. In addition, the Real Plan was introduced in the same year as the
Mexican Tequila crisis broke out in Latin America. This placed the entire strategy
in jeopardy and required a policy response of extremely high interest rates to pro-
tect Brazil from contagion and preserve capital inflows. The result was to place
further pressure on the fiscal deficit as the foreign balance continued to deteriorate
as the exchange rate continued to appreciate in real terms. The decision to intro-
duce a controlled nominal depreciation of the currency was finally taken in 1995,
but the continued decline in the rate of inflation was sufficiently rapid that the new
exchange rate regime gave little relief to exporters and the real appreciation appears
to have continued, although with a declining nominal rate.
Although the return of capital flows to Brazil in the 1990s started as short-term
speculative flows, the stabilization of the economy soon brought the return of
foreign direct investors. This reinforced the negative impact of liberalization on the
trade balance. Brazil had been a relatively closed economy, and as such most Brazil-
ian firms produced primarily for the domestic market. Thus, even the liberalization
of the import of capital goods by domestic firms did little to provide an offsetting
increase in exports, while the operations of transnational corporations involving
the final assembly of semi-finished goods all tended to increase the import content
of exports and raise the trade deficit that would be associated with a given growth
rate. Or, to put the point differently, the reduction in growth necessary to restore

438 Brazilian Journal of Political Economy 19 (3), 1999 • pp. 426-441

foreign balance was continually increasing over the period after the introduction
of the Real Plan. This is a response similar to that of Mexico, but without the
benefít of the linkage to the expanding US economy. Indeed, Brazil was enjoying
an internal consumption boom.
But there is another, often unnoticed impact of international financial market
integration and the use of high domestic interest rates to attract foreign investments.
Since increasing global capital flows meant an increasing proportion of Brazilian
public debt came to be owned directly or indirectly foreigners. Thus the rate of
interest effectively paid to foreign holders became the domestic Brazilian rate. Fur-
ther, increasing proportions of the debt were dollar indexed, but paid domestic
interest rates. The net result was that an increasing proportion of the domestic
Brazilian debt was indexed to the overnight domestic interest rate, and thus an
increasing proportion of the externally owned debt was also linked to the domestic
overnight rate. Just as profits on foreign direct investments, these interest earnings
appear as a debit item on the services, or liquid balance of payments. They represent
the same threat of instability as the unremitted profits of foreign direct investors.9.
Changes in domestic interest rates thus have an increasing impact on the current
account deficit as well as on the fiscal deficit and they may be expected to move in
tandem. This is a vicious circle that is similar to that caused by high interest rates
under the period of hyperinflation. To attract the capital inflows required to keep
the exchange on its target path and balance the fiscal and foreign deficits has re-
quired increases in interest rates that simply increase the size of these deficits. Fight-
ing inflation through exchange rate stability via increased interest rates has brought
price stability but at the cost of the foreign balance which eventually undermines
exchange rate stability. Indeed, this policy has not only immobilized the interest
rate as a policy tool, but it has also blunted fiscal policy as a result of the impact
of debt service on the deficit. Brazil is widely believed to have a fiscal crisis, but its
fiscal ratio is not far different from that of most developed countries and like Italy
it has had a primary surplus (although this has been due in some cases to the real-
ization of revenues from privatization) – the deficit is entirely composed of debt
service. Cuts in fiscal expenditures are required to offset the impact of interest
payments just as increasing exports are required to offset the effect of higher inter-
est rates on the services payments of the current account balance.
This scenario looks very much like that sketched out in the introduction. It
was the success of the policy that attracted the capital inflows, but this required
interest rate policies that produced deterioration in both the fiscal and foreign bal-
ances that eventually created doubts about the long-term success of the policy and
a reversal of capital flows. In difference from Asia, virtually all market participants
recognized that the policy was not tenable, and the capital outflow occurred slow-

9I have discussed the impact on financial instability of these flows in “Some Risks and Implications of
Financial Globalization for National Policy Autonomy”, UNCTAD Review 1996, March 1997.

Revista de Economia Política 19 (3), 1999 • pp. 426-441 439

ly, between July and January, rather than all at once. The result, however, was ex-
actly the same. A full scale exchange rate crisis.
Is this a crisis? Delfim Netto estimates that the current rate of growth of the
Brazilian labour force is around 2.7% per annum and that productivity growth is
around 3.5%, which means that Brazil requires a growth rate in excess of 6% to
keep the level of unemployment from rising. There seems to be no plausible way
the economy will be able to generate that rate of growth without drastically reduc-
ing interest rates. Just as Italy and the UK in 1992 were judged to have exchange
rate targets (and thus interest rate policies) that were incompatible with politically
acceptable rate of unemployment, Brazil was also so judged by international mar-
kets. Given the interest rates that prevailed in Brazil before the outbreak of the
crisis, the policy was already untenable, since export growth was not sufficient to
cover the increasing deficit on the services balance. The Asian crisis forced the issue
by pushing up interest rates to levels which produced visible changes in both the
foreign and fiscal balance and forced the introduction of fiscal austerity policies
which made the impact on employment growth clear and visible. The slow drain
of reserves and the slow flight of foreign investors10 meant that a meltdown of the
Asian style was avoided, but no amount of additional reserves in could change the
basic fact that the interest rate on most of the outstanding debt obligations was
higher than either the rate of growth of the economy or the rate of growth of ex-
ports, and thus neither the fiscal nor the foreign balance could improve without a
drastic fall in interest rates. The exchange rate thus had to be left to the market.
There are some who suggest that this will be achieved as Brazil produces a large
export surplus as a result of the devaluation. But, much like the initial Asian response
and given the Brazilian structure of trade after liberalization, the trade surplus is be-
ing produced through a fall in imports that is more rapid than the fall in exports, as
the sharp decline in incomes cuts consumption and imports. This suggests that a
decline in interest rates that lead to a resumption in growth high enough to reduce
unemployment is likely to recreate the foreign imbalance, requiring a return to high
interest rates and more fiscal austerity leading to another reduction in growth.

V. POLICY ALTERNATIVES

The real question facing Brazil in identifying an alternative strategy is what the
aim of policy is to be. Is it to borrow more from abroad by restoring confidence in
the currency and attracting foreign lenders or is it to earn more by attracting foreign
buyers for Brazilian output? Is the aim of fiscal policy to reduce the deficit by cut-
ting expenditures or by increasing the rate of growth of GDP. If it is to be the latter,
then the recent return to capital markets is not a sign of health, but of weakness.

10 US banks cut their exposure to $18.6 billion by the end of September, a reduction of 25% between
June and end September.

440 Brazilian Journal of Political Economy 19 (3), 1999 • pp. 426-441

The failure of exports to expand is not a sign of health, but of weakness, as is the
recovery of the exchange rate. If improvement is an increase in capital inflows, an
appreciation of the currency, but interest rates above 20%, then the Brazilian crisis,
following the path of Mexico, will not be the last.
In domestic capital markets, lenders are supposed to exercise diligence and
governance in order to insure that no borrower becomes excessively leveraged so
as to present a threat to the stability of the financing system. The example of LCTM
in the US suggests that even in the most advanced form of governance, supervision
and control, the market is not capable of imposing discipline intime to limit expo-
sures to safe levels. This has now been admitted even by those who champion the
minimum amount of regulation and maximum amount of freedom in capital mar-
kets (cf. President’s Working Group on Financial Markets, 1999: viii). Global mar-
ket mechanisms are even less well developed, which suggests that the responsibil-
ity for limiting leverage to that which is compatible with domestic economic
stability must lie with national governments. This implies some for of rationing of
capital inflows when they exceed that which an economy can safely absorb in the
context of its domestic stability policies. Given the strong qualifications required
for the validity of the Modigliani-Miller theorem (which says that capital structure
has no impact on a firm’s performance) finance economists have investigated the
optimal degree of leverage for an individual firm. Perhaps it would be more ap-
propriate if instead of discussing the control of capital flows the question were
posed in terms of the optimal degree of leverage for a developing country.

REFERENCES

DELFIM NETTO, Antonio (1998). “O desemprego é a âncora do Real”, in Visões da Crise. Rio: Con-
traponto.
KREGEL, J. A. (1994-95). “The Viability of Economic Policy and the Priorities of Economic Policy”,
Journal of Post Keynesian Economics, Winter, 17: 2, pp. 261-77.
KREGEL, J. A. (1997). “Some Risks and Implications of Financial Globalization for National Policy
Autonomy”, UNCTAD Review 1996, March.
KREGEL, J. A. (1998). “East Asia is not Mexico: The Difference between Balance of Payments Crises
and Debt Deflation”, in JOMO, K. S. (ed.), Tigers in Trouble. London, Zed Books.
KREGEL, J. A. (1998). “Yes, ‘It’ Did Happen Again – A Minsky Crisis Happened in Asia”, Jerome Levy
Economics Institute Working Paper # 234.
OHNO, Kenichi. “Exchange Rate Management in Developing Asia: Reassessment of the Pre-crisis Soft
Dollar Zone”. Mimeo, Asian Development Bank Institute (nd.)
SUMMERS, Lawrence (1998). US Government Press Release, 2286, March 9: “Deputy Secretary Sum-
mers Remarks Before The International Monetary Fund”.
UNCTAD (1995). “Trade Development Report”. Geneva: United Nations.
US GOVERNMENT (1999). “Report of the President’s Working Group on Financial Markets: Hedge
Funds, Leverage and the Lessons of Long-Term Capital Management”. Washington, D.C. April.

Revista de Economia Política 19 (3), 1999 • pp. 426-441 441