The China Syndrome or The Tequila Crisis
Documento publicado con ligeros cambios en Latin America Macroeconomic Reform: The Second
Stage. Edited by Anne Krueger, Jose Antonio Gonzalez, Vittorio Corbo and Aaron
Tornell. The University of Chicago Press 2003. Con el título THE CHINA
SYNDROME OR THE TEQUILA CRISIS
There have been 112 worldwide banking crises in just the past 30 years,
three of them in Mexico (1977, 1982 and 1995), comprising 93 countries (Sánchez
Santiago 6/6/00 and Klingebiel 2000). These economic breakdowns have frequently
been followed by salvage operations of bank depositors that involve
considerable jumps in public debt. Besides the fiscal costs, there have been
substantial capricious wealth transfers, personal and corporate bankruptcies,
output contractions, inflation, and often a severe questioning of the efficacy
of the market model for economic organization. A cause and effect confusion
about the causes of some of these crises has led even reputable economists to
question some market features, such as freedom of capital movements, at least
as temporary measures for some economies (Krugman, 1998 and 1999). Given the
importance of the matter and the lack of consensus between economists and/or
policy makers about the origins of these crises, a search for answers to this
transcendental matter is imperative. Perhaps the fundamental question ought to
be: have we faced policy or institutional failures?
As with many
matters, the answer may have eluded us because we have failed to pose the
proper questions. Money meltdowns (Shelton, 1994), whether in the form of
national banking collapses or international epidemic (systemic?) crises, are
attributed to a wide diversity of causes, but seldom (Hayek, ) to fundamental institutional flaws. Insufficient
regulation and supervision of financial sectors, the popular whipping horse,
may even be a source of moral hazard. Therefore, when dealing with regulation
and supervision, it might prove fruitful to attempt to identify the ultimate
causes of moral hazard. At the macro level the primary source of financial
shocks seems to be an erroneous choice of the exchange rate regime. At the
original sin level (Hausmann and Eichengreen, 1999, apply this term to another
phenomenon) the ultimate root of moral hazard is central bank credit. On this
epistemological vein, exchange rate “regimes” owe their existence to central
banks.
The large
number of financial crises of recent, their substantial contagion effects onto
emerging markets, because the most recent viruses (Mexico, Southeast Asia,
Russia) emerged from them, and their potential for creating destructive chain
reactions even among developed markets have given rise to renewed research
about their causes. The belief that lack of supervision of financial
institutions is the primary cause has spawned proposals for a cosmopolitan
approach to supervision and considerable thinking is being dedicated to the
design of new rules, with concentration on improving opaque and untimely
information from some emerging markets.
The reason for
the uniqueness of financial markets in this regard is the potential for
fraudulent or careless behavior (moral hazard). Fraudulent potential comes of
course with every contract, but systemic fraudulent potential has an ulterior
root: central banks. In any national economy the singularity of central banks
originates in their ability to create unlimited quantities of domestic base
money. In turn, when faced with liquidity problems national governments turn to
the international community for international base money. The possibility of
recourse to outside support is a source of moral hazard, but the potential for
abuse is enhanced because assistance does not have to come from real resources
(savings), some central banks and the IMF (SDR’s) have the potential to issue
internationally accepted means of payment[1]. These
international lines of credit are drawn upon sometimes automatically and often
as a result of concerted international rescue efforts. Therefore, national
moral hazard is compounded by international moral hazard and may ultimately
even be the result of the latter.
The fact that
means of payment are connected, sometimes in seemingly interminable chains that
would collapse and irradiate multiple reactions if one of their links were
destroyed, is also a cause for concern and may even have originated the
perceived need to have a lender of last resort. There is a generalized opinion
that payment chains cannot be broken, that in a fix liquidity has to be
provided at all costs. Under this assumption we have an explanation for the need
for national and world central banks. Without a gold standard another
justification for the need of central banks are growing economies: if the
fiduciary money required by the growth of transactions is not provided, the
price level would have to fall continuously.
Audit rules,
deposit insurance, limits on access to last resort lending, capitalization
requirements, liquidity ratios, asset “mark to market” valuations, accounting
standards, rules of disclosure, “fit and proper” controlling shareholders and
administrators are requirements designed to contain the vast centrifugal forces
unleashed by the incentives to abuse public trust. Most of these rules attempt
to solve the problem by attacking the root problem: the prodigal uses of base
money, and could even end up abolishing supervision.
The causal
chain is thus originated at central banking. Central banking is at the source
of potential unlimited lending of last resort; in turn commercial fractional
reserve banking creates a daily need for such (intra-day or end-of-day lending)[2] last
resort lending, but also enhances the quantities that may eventually be
required from the central bank. Fractional reserve banking also increases the
asymmetry in returns that may be obtained from speculative or fraudulent
banking investments: owners or administrators will benefit from leveraged
investments but may bear only a fraction of the losses.
The
monetary policy origins of moral hazard
First, as
promised, some comments on the implications for moral hazard of exchange rates
regimes. Despite all the hoopla about exchange rate combinations, there are
only two possible regimes. A fixed exchange rate at one extreme, and a freely
floating one at the other. Combinations converge with time into the first
regime. A lower or an upper exchange rate band, or both, as departures from the
two extreme regimes to achieve some eclectic combination, implacably end up
with the characteristics and vices of a fixed regime. The implications of a
fixed regime for moral hazard are clear: it places the central bank at the
mercy of short-term international money managers. At one of the two extreme
possibilities for the exchange rate, an excess supply of foreign currency tends
to appreciate it. To hold on to the exchange rate, the central bank will have
to purchase the excess supply of foreign currency, will accumulate reserves,
will issue internal domestic debt and will end up subtracting credit from the
economy.[3]
All recent
international economic crises[4] have a
common element: a commitment to defend fixed nominal exchange rates or exchange
rate bands that coexisted with substantial capital inflows, i.e., large current
account deficits. The aftermath of these crises also had a common outcome: a
subsequent banking collapse and rescue of depositors. These experiences raise
the customary chicken and egg question; did “good” economic policies attract
net capital inflows? Or, did fixed nominal exchange rates because of the high
yield implicit in one-sided bets lure “excessive” net capital inflows that were
not necessarily well invested? The experiences of Mexico[5] in this
regard prior and after the crisis of December 1994 are telling. Short-term
capital flows were large and volatile during 1994 leading to the Dec. 1994
crack, while becoming small and stable since the beginning of 1996. The
explanation to such a behavior reversal may lie in the moral hazard originated
in an exchange rate commitment, whose effects are accentuated by the speed and
liquidity that characterize today’s financial markets.
Consider the
amounts invested by foreign residents in Mexico’s money market (purchases by
foreign residents of government securities and other money market instruments)
as of two significant dates: end of Dec. 1995 and end of Dec. 1999[6]. The
period chosen is significant because prior to December 1995 the figures are
contaminated by the liquidation of Tesobonos by the Mexican government, the now
infamous dollar-linked peso securities first issued at the end of 1991. At the
end of 1995, the stock amounts, not the flow, of these investments was $3.8
billion. One year later, on Dec. 1996, it had barely edged up to $3.9 billion
and to 4.1 billion by the end of 1999. Hence over the course of three years the
figure has stayed basically the same. Not only was the change in the number
minimal, the ups and downs of this concept over the 36 month span were also
insignificant.
The immediate
post-crisis period, that is to say 1995, is atypical, because during that year
the Mexican government decided to stop issuing Tesobonos and to liquidate those
outstanding. The behavior of short-term capital from December 1995 up to the
present has already been detailed above. The turbulence of that year’
continuous crises, including some political ones, also justify excluding it from
the comparison.
The behavior
described above is more remarkable given the large influx of other categories
of foreign resident capital into Mexico: during the past three years the
country has been the second largest emerging market recipient of foreign direct
investment in the world, second only to China. It has also received large
equity investments as well as a resumption of foreign bank loans and the
floating of private liabilities in international money markets. These flows
contributed to finance current account deficits that cumulated $ 39.5 billion,
plus an increase of international reserves of $ 18.2 billion.
Not only did
the amount of foreign resident’s money channeled to Mexican monetary
instruments remain stationary over this latter period, but it is also more
resilient to shocks: its term to maturity is considerably longer than the
forty-eight-hour investments that flowed in large quantities prior to the 1995
crisis. The instruments now being purchased by foreign residents tend to have a
maturity of at least three months, with six months to one year being the
favorites.
Along this line
of reasoning Trigueros (1997) concludes that foreign direct investment,
portfolio investment and foreign currency deposits issued by commercial banks,
as well as the direct credit they obtained, exhibited remarkable stability
after the onset of the crisis. The opposite is true of foreign exchange inflows
to the Mexican money market prior to the crisis (prior to the flexible exchange
rate), when from 1990 to September 1994 $40 billion in short-term capital
poured in.
The shift in
the exchange rate regime between the pre-crisis and post-crisis periods may
explain the radical change in the nature of foreign capital flows into Mexico.
The current stability of short-term capital flows, to the extent that their
variability reflects the volatility of the stock demand, may hold the key to
understand the remarkable stability of the exchange rate over the recent
period. In turn, the perilous situation into which the Mexican economy fell
before 1995 was the outcome of pathological market behaviors provoked by the
fixed, or quasi-fixed, exchange rate that prevailed at the time, without the
automatic self-adjusting processes of a currency board.
During the
years under the peso-dollar band (1990 ?- 1994), when the exchange rate veered toward the floor of
its initially narrow interval, a combination of high peso yields and exchange
rate appreciation attracted large volumes of short-term capital inflows. Once
the exchange rate stuck to its floor, investors would continue to obtain high
yields under the understanding that the exceptional returns were possibly
transitory, because the availability of hard currency, despite the implicit
promise of convertibility, had as a limit a fraction of the country’s
international reserves. The interest rate required to bring capital in, given
those uncertainties, had therefore to include a premium, and the term for which
money market investors were willing to commit their capital had to be extremely
short, allowing them to keep one foot in and the other outside, so to speak.
As long the
central bank had sufficient international reserves the other extreme
possibility was for the exchange rate to be at its peso-dollar ceiling. During
the existence of the band, this phenomenon only occurred when radical political
events[7] created
confidence shocks. Each of those shocks drained international reserves. At
those junctures the incentive for investors was to try to be the first out of
the local currency, in a situation that corralled them into such a desperate
corner that virtually no interest rate would have been sufficient to encourage
them to stay.
Mexico’s
experience validates once more that an exchange rate commitment generates two
polar possibilities. One is to attract vast foreign inflows invested in
extremely short-term instruments, the other, to have investors fly away as
quickly as possible. But this is old stuff, of course: that exchange rates tend
to veer off to their allowable extremes, and that, once there, the system has
all the flaws and dangers of a fixed exchange rate, was pointed out several
decades ago by Harry Johnson. Otherwise, if the exchange rate does not stick to
either of its extremes, the system behaves as a floating exchange rate. So why
contaminate it with bands? Unfortunately, it seems necessary to continue
rebottling old wines for the consumption of some economists, as well as for
policymakers.
To conclude
this section, the case study of Mexico is congruent with the hypothesis that
one of the sources of moral hazard, in this case with substantial macroeconomic
consequences, is the incentive provided to speculators to play the short-term
capital Ponzi scheme created by fixed exchange rates. This behavior adds to
whatever problems the commercial banking system may face from its own moral
hazard roots and may be sufficient to create a banking crisis, without having
government officers and banks engage in irresponsible acts. But the exchange
rate regime was not the only economic policy flaw.
The Banking Original Sin (Or Home Made Moral Hazard)
With some
important exceptions (Hayekians[8] and
followers of Henry Simons among others), economists tend to confuse bank demand
deposits[9] with
savings to be used for long-term purposes. A representative example from the
literature is the following: “banks issuing demand deposits can improve on a
competitive market by providing better risk-sharing among people who need to
consume at different random times”. (Diamond and Dybvig, 1983. Reprinted in
Federal Reserve Bank of Minneapolis, 2000).
In a Fisherian
intertemporal model (Fisher, The Theory of Interest, 1930) there is room for
intermediary institutions (Mackinnon,_____) that select borrowers (or claims issued by negative
savers) and diminish portfolio risk by bundling assets with negative risk
covariance, and that in turn issue liabilities held by individual savers. But
there is no reason for these institutions to be also issuers of demand
liabilities, they can issue consols or time matched claims on themselves. In
fact, whatever the time distribution of the consumption pattern of savers,
demand deposits run the risk of being massively and instantly converted into
cash and may lead to an unmanageable run with systemic and macroeconomic consequences.
Given these risks, the only possible explanation for institutions with grossly
mismatched balance sheets is the existence of someone willing to come to their
rescue in case of need. Such is the source of original sin, which of course
lays the foundations for the committing of venial sins such as credit to the
private sector financed with sight deposits. Venial sins are transformed into
mortal ones when the size and growth of such credit become excessive.
A recent
example of credit expansion going haywire is Mexico where credit growth prior
to the crisis was astronomical: 25 percent per year in real terms for six
consecutive years. A substantial portion of the loans that the banking system
churned out in this interval had a poor or nonexistent possibility of recovery
even before the skyrocketing interest rates and the exchange rate depreciation
that ensued December 1994.
Several factors
enhanced the intrinsic deposit banking moral hazard responsible for this
expansion: a) after several years of government ownership the human capital of
commercial banks had eroded considerably, b) the capitalization of some banks
was thin or completely transparent, and c) the financial system underwent a
substantial liberalization.
The
liberalization of the banking system released sudden copious resources that
banks felt compelled to lend, the increase in lending induced an increase in
aggregate demand that would in turn widen quickly and excessively the deficit
on the current account of the balance of payments. The new liberal measures
included the disappearance of forced sector loans, the freeing of interest
rates, and the elimination of reserve requirements.
Among recent
crises the Mexican experience is not unique. We know now that collapsed Asian
countries also experienced vast credit expansions of dubious quality. The
similarities between the Mexican and the Asian crises and others include as
well astronomical increases in real estate prices[10]. More
fundamentally, because of their exchange rate systems, excessive amounts of
short-term money[11]
were fatally attracted. The similarities end there, however, because Asian
countries’ high export growth had petered out prior to their crises, while
Mexico’s non-oil exports were growing in 1994 at a pace of 20 per cent, over an
already high base[12].
As we realize
the nature of Mexico’s and of other recent crises, one of the fundamental
questions raised is, what should economists watch? Reflecting on the evidence
reviewed and on recent developments[13] it is
becoming ever more evident that economists should discard the Freudian (anal?)
obsession with the real exchange rate, or with the current account of the
balance of payments. But as we shall see, perhaps a combination of the
(well-measured) public budget, the growth in credit and some market oriented
measurement parameters of the health of the financial system are the symptoms
one should be aware of.
Another issue
raised by recent crises is the appropriate exchange rate regime. Some
economists or policymakers may opt for what I believe are the increasingly
futile bands, others for flexible exchange rates. Experience suggests that the
answer is to be found at either of two extremes: either no autonomous issuing
of currency at all -with the currency board as an approximation to this
solution- or a flexible exchange rate. However, a well-functioning flexible
exchange rate requires institutional buttress, is not sufficient to simply let
it loose. Among other ingredients coverage mechanisms are essential, local
institutions that allow for cover may be found wanting, however. At the end of
the leg of a transaction investors have to be assured of the delivery of hard
currency. The importance of this ingredient was evident at the outset of the
Mexican crisis. Despite a deep local market in forward contracts, the foreign
exchange market did not contribute to the stabilization of the peso and of
local interest rates until the appearance of futures transactions in the
Chicago Merc guaranteed delivery[14].
Having expanded
on the pros and cons of alternative exchange rate regimes and their
implications for the attraction of some particular capital flows, one should
pay tribute to the obvious, to the other institutional elements essential to a
well functioning economy and therefore to a well behaved floating rate. Recent
currency stability in Mexico was aided by the reforms detailed above, but it
would not have been possible without deep adjustments to the government budget,
without wage revisions that have not outstripped gains in productivity, nor
without political stability.
The next few
paragraphs will be dedicated to identify the triggers of the 1995 economic
implosion under a quasi-fixed exchange rate. One frequently cited is the Fed
tightening of 1994. Fed tightening does matter; in the early 1980s Fed actions
were an important part of the explanation of international economic behavior.
Rates went up dramatically in the US and worldwide. They had a big impact on
the US economy and the economic slowdown depressed Latin American export
prices. But did the 1994 jacking up of US interest rates really matter that
much to Mexico, or did the political shocks of that year trigger the crisis,
together with the fact that Mexico’s current account was getting far out of
line given banking mismatches and the overhang of a huge amount of liquid
liabilities redeemable in dollars under declining international reserves?
How important
are those retrospectively relatively small movements of US interest rates in
1994 as a driver of what happened to the peso/dollar exchange rate? In the
context of the accumulated pressures and the knife-edge the economy was
traversing in 1994, the US interest rise contributed to aggravate Mexico’s
travails but it cannot be singled out as the cause of the crisis.
The answer
should be sought in the pressures the economy had accumulated like a pressure
cooker during the former 4 or 5 years. Said pressures are closely related to
banks and to their privatization. Despite some common elements not all bank
privatizations were flawed. There were three kinds of banks, or three groups or
tiers of purchasers of banks. One tier of shareholders immediately started
figuring out how to conduct fraudulent operations with their banks. It is not a
question of bad loans; it is a question of black holes in their accounts, money
whose final destination has not been found. The amounts irretrievable climb can
go up to several billion dollars when looking at those operations. Some of the
large irrecoverable amounts, totaling several billion dollars, were not an
isolated case of maybe a bad loan or a sour operation. Some of the small banks
or medium-sized banks that comprised this tier were in the hands of people who
are now being prosecuted or who are abroad facing extradition by the Mexican
government. Most of these funds were simply channeled to their owners’ private
uses.
The next tier
raised the money needed to bid for the banks by convincing potential investors
that they would enjoy large capital gains and promising to issue them loans to
pay for the shares they were committing to buy as soon as they had control of
the bank. The first tier also used this capitalization ruse, but it was not the
root of their failures. Regarding this second tier, bidders paid too little or
nothing, a consolidation of assets and liabilities would have cancelled most of
their apparent capital. In this tier and in the other 2 the capital of some
banks was subscribed with “cross-financing” that involved an understanding
between different groups of bidders to lend from bank A to the purchasers of
bank B in order to have bank B do the same with the purchasers of bank A. There
were even instances of development bank lending to buttress some acquisitions.
The third tier
is where the two big banks are situated, as well as a few others, some very
small ones actually (Banorte is the prime example). In this tier one finds that
real capital was put in and that there were no fraudulent schemes; but even in
these cases there was some poor lending, although most of this bad lending had
taken place before the banks were privatized.
At the
beginning of the Salinas presidential term commercial banks found themselves
with a sudden gush of funds while still in government hands. The reason is that
there was a reduction in government debt. Internal Government debt went down
from 20% of GDP to 5% of GDP over a couple of years as the proceeds of the
privatization of all kinds of firms went, as well as some government surpluses,
into a liquidation of that debt. Besides, the successful renegotiation of the
government’s external debt was also concluded before the banks were privatized,
putting Mexican banks again into the international borrowing market. These 2
developments propitiated rapid credit increases, and, under weak management and
supervision, the bad loans started to be generated.
A Leveraged Buy Out, Complementary Regulatory Failures
and the Onset of the Crisis.
The financial sector liberalization
process was far reaching. It introduced several desirable features that should
provide for a more competitive banking sector. But as will be seen throughout
this section, the reform was incomplete; some of its features encouraged an
increase in the supply of credit of such magnitude and speed, that it
overwhelmed weak supervisors, the scant capital of the banks and even
borrowers.
Besides the 2 elements described
above, several other factors contributed to facilitate the 1988-94-credit
explosion: improved economic expectations, a real estate and stock market [15] boom
and a strong private investment response. The latter due in part to the need to
adjust the capital stock to the substantial restructuring requirements that the
trade opening and market deregulation placed on the economy. There was an
abundance of loanable funds because of the reduction of public debt and the
phenomenal increase in the domestic and foreign availability of securitized
debt.[16]
Table 1
Price Index of
Urban Land in Mexico City
(End of period,
1980 = 100)
Year
|
Price Index
|
1987-IV
|
1,930.29
|
1988-I
|
2,422.87
|
1989-I
|
7,814.56
|
1990-I
|
12,012.23
|
1991-I
|
16,451.04
|
1992-I
|
23,079.09
|
1993-I
|
31,933.75
|
1994-IV
|
33,927.21
|
Source: Banco de Mexico. Dirección General de Investigación Económica.
A balance sheet adjustment of the
private sector underway by the second half of 1993 and the late adoption by
some commercial banks of more prudent policies were signs that the
non-performing loan problem had exceeded manageable dimensions before 1994:
“Wide insufficiency of capital was becoming perceptible, a phenomenon explained
by the relatively high level of past due loans that had not been adequately
provisioned. Moreover, some commercial banks were operating with serious
problems that were not readily noticeable to the financial authorities. In some
instances, bank administrators acted with disregard to existing regulations and
proper banking standards.” [17]Consumer
and housing credit increased because of natural causes but banks pushed even
these loans beyond reasonable standards. There was poor borrower screening and
credit volume excesses, and then the 1993 economic growth slowdown made the net
indebtedness of the private sector burdensome.
As of the second quarter of 1994
sharply higher real interest rates [18] and a
considerable but still orderly depreciation of the peso, prior to the December debacle,
the aftermath of the assassination of a presidential candidate and other
unfortunate political events, poured gasoline onto an already burning coal.
The devaluation of December 1994 had
a limited direct impact on the financial position of commercial banks.[19]
Nonetheless, the devaluation prompted other damaging effects as inflation and
interest rates skyrocketed, economic activity collapsed, the burden of
servicing credits denominated in domestic and foreign currency increased, and
banks’ capitalization ratios collapsed.
The financial situation of private
firms evolved from an unprecedented net asset position at the end of 1988, to a
substantial, quick and fragile indebtedness by 1994. A remarkable coincidence
of converging events and policy measures combined to produce this phenomenon.
In fact, all the cards were stacked in favor of a collapse.
a) The financial sector was
liberalized, lending and borrowing rates were freed and some remaining credit
channeling was abolished.
b) Government
surpluses drastically reduced public internal peso debt. The debt reduction
freed the portfolio of banks and facilitated the elimination of reserve
requirements.
c) To calculate
non-performing loans, banks applied a “due payments criteria”: the amount of
payments due after 90 days was recorded as delinquent, instead of the value of
the loans themselves.
d) Banks were
hastily privatized, in many instances with no due respect to “fit and proper”
criteria in the selection of shareholders, nor of their top officers.[20] It must
be noted, however, that the banks remained in government hands for half of the
expansionary period and that part of the sour loans had already been extended.
(Table 2)
Table 2.
Privatization Dates of Commercial Banks
Bank
|
Date
|
Multibanco
Mercantil de México
|
August, 1991
|
Banpaís
|
”
|
Cremi
|
”
|
Confía
|
September,
1991
|
Banco de
Oriente
|
”
|
Bancrecer
|
”
|
Banamex
|
”
|
Bancomer
|
October, 1991
|
BCH
|
November,
1991
|
Serfin
|
January, 1992
|
Comermex
|
February,
1992
|
Somex
|
March, 1992
|
Atlántico
|
”
|
Promex
|
April, 1992
|
Banoro
|
”
|
Banorte
|
June, 1992
|
Internacional
|
”
|
Banco del
Centro
|
July, 1992
|
Source: Taken from Guillermo Ortiz-Martínez (1994). La Reforma Financiera y la Desincorporación Bancaria. México D.F.
Fondo de Cultura Económica.
e) Several
banks were purchased without their new owners proceeding to their proper
capitalization, as required by their financial situation, since shareholders
often leveraged their stock acquisitions, sometimes with loans provided by the
very banks bought out or from other reciprocally collaborating institutions.
f) Taxes on
inter country capital flows (dividends, interest, etc.) were drastically
reduced or eliminated.
g) Foreigners
were allowed to hold short-term “domestic” government debt as of December 1990.
h) Short-term,
dollar-indexed, peso-denominated Mexican government securities, Tesobonos, were
issued at the end of 1991.
i) The higher
echelons of banks lost a substantial amount of human capital during their
government years. With these officers institutional memory migrated as well.
This experience is not unique to Mexico: “...formerly regulated banks may lack
the necessary credit evaluation skills to use newly available resources
effectively.” [21]
and: “Unless properly overseen, liberalization can result in too rapid growth
of bank assets, over-indebtedness and price-asset bubbles.” [22]
j) Full
government backing of bank deposits.
k) There were
no capitalization rules based on portfolio market risk. This regulatory failure
encouraged asset-liability mismatches that in turn led to a highly liquid
liability structure, more than two thirds overnight for the banks, with a
potential to create huge strains on the lender of last resort capabilities of
the central bank.
l) Banking
supervision was weak and also overwhelmed by the great increase in the
portfolio of banks.
m) Some
commercial banks abused the unlimited supply of daylight overdraft facilities
at the central bank and created vast amounts of deposits drawing on the inter
bank credit market and on the money market.
n) The banking sector did
not have a consumer loan credit bureau nor did it actively utilize the business
bureau available.
o) The wide
trade opening and deregulation that swept the economy altered relative prices
and canceled opportunities in traditional sectors. Formerly privately “good”
projects turned into bad ones and altered the relative ability to service debt
of many sectors and or types of enterprises.[23]
p) There was a
phenomenal expansion of credit from the development banks.
q)
Unprecedented and huge amounts of foreign capital became available worldwide
and particularly to Mexico. One of the salient newcomers to these capital
markets were securitized flows.
r) The banking sector faced
procedural and judicial difficulties that enhance the spread between lending
and borrowing rates. This problem was considerably accentuated by the onset of
inflation (in the former inflationary period banks did not have a substantial
amount of credit outstanding) and the economic collapse of 1995.
Table 3
Dollar Denominated
Financial Obligations
Of the Financial System
November, 1994
|
US Dollars
|
Balance of
the External Debt of the Commercial Banks
|
25,966.0
|
Balance of
the Short Term Foreign Debt of the Development Banks
|
4,562.2
|
Tesobonos
|
24,690.7
|
Total
liabilities of the Commercial Banks with Mexican Residents (net of inter bank
operations)
|
112,902.0
|
Net
International Reserves
|
12,483.9
|
Cash in
circulation
|
14,251.1
|
1995
|
|
Interest on
the Foreign Debt
|
11,715.9
|
Public Sector
and Banco de México
Private
Sector
|
7,368.9
4,347.0
|
All the
elements listed above combined with a greatly improved perception of the
country’s short and long-term prospects to generate the conditions that would
result in the Mexican crash. A paragraph from Linden et. al. portrays well what went on: [Referring to Mexico 1994-present] “After many years of nationalized banking [from 1982 to mid-1992], commercial banks lacked the experience and organizational and
information systems to adequately assess credit and other market risks and to
monitor and collect loans. Accounting practices did not follow international
standards. Concentration of loans and loans to related parties was a problem in
those banks that were subsequently subject to intervention.” [24] In
relation to many countries the Bank Soundness book finds that: “Banks
that are, or were recently, state-owned were a factor in most of the instances
of unsoundness in the sample” [25] and “It
becomes more difficult to distinguish good from bad borrowers when bank loans
are growing rapidly.” [26]
This combination of factors
constitutes another experience [27] of how,
despite important economic achievements, financial liberalization [28] can go
stray in an environment that has no adequate safeguards against the predatory
practices banks can be induced to by full deposit protection. Thin or no
capitalization was another key ingredient that combined with the other factors
to induce imprudent credit growth.[29]
Fast credit growth and its aftermath
is not an exclusive feature of Mexico’s crisis.
Chile’s 1982 crisis suffered from the same and Kaminski and Reinhart
(1996) reviewed the experiences of 20 countries that experienced banking and
balance of payments crises and found that in about half, the banking crisis
preceded the balance of payments crisis. The causal pattern reversed in only a
few instances. Thus, there is support for the notion that bank soundness exerts
negative effects on the external balance and the exchange rate.” [30] Also:
“All the sampled countries except Venezuela experienced a sharp expansion of
credit to the private sector prior to the crisis.” [31]
The numbers related to the expansion
of credit during the Salinas administration are impressive. From December 1988
to November 1994, the amount of credit outstanding from local commercial banks
to the private sector rose in real terms from 90.3 billion pesos to 340
billion. Also in real terms, the relative increase in this credit over the
six-year period was 277 per cent, or 25 per cent per year. [32]
Some items of this expansion provide
a better picture of the trends that characterized it: credit card liabilities
rose 31 per cent per year, direct credit for consumer durables rose at a yearly
rate of 67 per cent and mortgage loans at an annual rate of 47 per cent. All
these rates of growth are in real terms.
In dollars, external credit flows to
the private sector went from -193 million in 1988 to 23.2 billion in 1993 and
to 27.8 billion in 1994. The flow fell to 8.9 billion in 1994, but this
decrease was more than compensated by the lower international reserves of Banco
de México that year, which went down by 18.9 billion. Therefore, the total use
of external resources was of 27.8 billion in 1994.
The accumulated amounts related to
external flows are also substantial.[33] A total
of 97 billion dollars over the six-year term that increases to 115.9 billion
once the fall in reserves that occurred in 1994 is included.
These rates of growth are portentous. As Honohan (1996 p. 1 of Annex)
warns, “...there are general indicators which apply whether or not there is a
macroeconomic boom and bust cycle.” Then Honohan lists among others, the
following telltale signs:
·
“One measuring balance sheet change, namely the growth
in aggregate lending (in real terms). This is the classic indicator of
individual bank failure and may also serve for systems.”
·
“Two drawn from the structure of the balance sheet,
namely the loan-deposit-ratio and reliance on foreign borrowing.”
Table 4
External
Financial Flows to the Private Sector
(Millions of
dollars)
Concept
|
1988
|
1989-1994
|
Total
|
943
|
97,096
|
Loans
|
-1,548
|
23,984
|
Banks
|
1,380
|
16,209
|
Non-banks
|
-2,928
|
7,775
|
Portfolio
|
-389
|
43,787
|
Shares
|
0
|
28,403
|
Bonds
|
-389
|
14,381
|
Direct Investment
|
2,880
|
30,325
|
Source. Balance of Payments Statistics. Banco de
México.
But the story is not yet complete.
Starting in 1993, the government
decided to break with a long and healthy practice of including in the
definition of its consolidated deficit the amounts channeled through government
development banks, a concept known as the deficit or surplus, as the case might
be, due to “Financial Intermediation”.
Financial Intermediation had been included in the deficit: a) to
restrain overall budget expansion, b) to prevent the use of development banks
to disguise public expenditures, c) because they expand credit based on central
and not market decisions, and d) their loans to the private sector were of
dismal quality.
Table 5
Deficit Due to Financial Intermediation
Year
|
% of GDP
|
1990
|
1.07
|
1991
|
2.80
|
1992
|
2.66
|
1993
|
3.33
|
1994
|
3.68
|
Source: Banco de México.
The abandonment
of the sound practice of including this deficit within the overall picture
contributed to an additional expansion of credit and to careless lending:
“During the past government Nacional
Financiera extended 470,000 credits, of which half were not viable...” and
“...they were not viable even before the crisis.” [34]
From the figures shown in Table 5 it
can be appreciated that the pressure from financial intermediation to GDP was
not negligible to explain the size of current account deficits that were 3.0
per cent, 5.1 per cent, 7.4 per cent, 6.5 per cent and 7.9 per cent of GDP each
year between 1991 and 1994. [35]
The unseemly attraction of foreign
resources, the liquidation of large amounts of government debt which crowded-in
low quality bank lending, combined with the moral hazard cocktail concocted by
the various measures already enumerated, nurtured an increase in private
aggregate demand which contributed to the rapidly rising current account deficit.
Furthermore, the deficit was financed in a large proportion by short-term
capital. This deficit was combined with the commitment, a pledge consecrated in
the Pacts, to contain the exchange rate within a widening, but relatively tight
band.
For most of the period the exchange rate
stuck to its peso/dollar floor, as high interest rates attracted short-term
capital, development banks and private firms borrowed abroad and foreign money
flowed into the stock market. The central bank essentially accommodated the
demand for base money (i.e., currency) and in that endeavor sterilized foreign
exchange inflows or outflows, allowing international reserve increases or
decreases as the case might have been. Because of the predominant excess supply
of dollars, the amount of reserves constantly increased, up to the uncertain
period prior to the US Congress vote on NAFTA, when the increase was
temporarily interrupted, only to resume after NAFTA was approved up to the
start of 1994’s political wobbles.
Table 6
International Reserves
(Billions of dollars at the end of
each year)
1988
|
1989
|
1990
|
1991
|
1992
|
1993
|
1994
|
6.4
|
6.6
|
10.2
|
17.5
|
18.6
|
24.5
|
6.1
|
Source: Banco
de México.
The deficit in the balance of trade rose 5.83
percentage points of GDP over the period, explainable up to 81 per cent [36] by the
rise in private investment. But a substantial portion of this increased private
investment went into unprofitable ventures, thus contributing to the
non-sustainability of the current account deficit. Some of these undertakings
were highly leveraged toll-roads, or unrecoverable home mortgages, or credit
unions that invested with low or negative returns and were financed with vast
resources channeled through the development banks. Some of the credit, in turn,
went to finance non-existent enterprises, or the hugely levered acquisition of
bank shares, or to non-collateralized loans, etc.
Table 7
Current Account Deficit as a Proportion of GDP
(Percentages)
1988
|
1989
|
1990
|
1991
|
1992
|
1993
|
1994
|
- 1.4
|
- 2.8
|
- 3.0
|
- 5.1
|
- 7.4
|
- 6.5
|
- 7.9
|
Source: Banco de México, Annual Report (1995).
Thus financial disequilibria, a classical overindulgence in credit, a
frenzy of spending and a substantial short-term debt combined with the
sitting-duck features of a fixed exchange rate, linked-up to set the stage for
the initiation of the 1995 economic crisis.
The crisis had little or nothing to
do with lower savings, as can be seen in Table 8, as argued by many, but a lot
to do with excessively rapid spending and credit expansion, as McKinnon and
Pill (1995), Calvo and Mendoza (1995) and Hale (1995) have pointed out.
Table 8
Consumption and Investment Growth
(Changes as a proportion of GDP in constant prices)
Consumption
|
Investment
|
|||||||
Years
|
X - M
|
Total
|
Private
|
Gov.
|
Total
|
Private
|
Gov.
|
|
1989-94
|
5.83
|
1.4
|
1.76
|
- 0.36
|
4.44
|
4.74
|
- 0.30
|
|
Source: Banco de México, taken from the National Accounts.
The virtually fixed exchange rate
exhibited its virtues by steadily stabilizing prices, as well as its dangers
within the environment created, particularly the fragility of the economy to a
speculative attack.
Just as many European currencies
collapsed in 1992 after unrelenting speculative attacks on their narrow bands,
1994 political events triggered what for one economist was a death foretold [37] but a
surprise nonetheless: a drain in international reserves until the exchange rate
ceiling had to be abandoned on December 1994.
A Financial Interpretation of the Crisis.
Are recent economic crises, the
European in 1992 and Mexico’s in 1994 the result of unsustainable policies
given unexpected shocks, or a reflection of multiple equilibriums not closely
related to measured fundamentals? [38]
The classic position related to
misaligned fundamentals can be traced back to Johnson (1972), where an excess
credit expansion is translated into a loss of international reserves. A balance
of payments crisis is the outcome of the depletion of international reserves.
This position can also be found in Sargent and Wallace (1981) who provide a
closed economy vision in which a persistent deficit and real interest rates
above the rate of economic growth eventually unclench debt saturation. At this
point private agents refuse to continue purchasing debt and the deficit is
monetized. Inflation ensues. This chain of events is not very different from
the open economy model. Finally, Krugman (1979) in a model reminiscent of
Mundell (1968) follows on this tradition in a futile attempt to time the
speculative attack, which will force an abandonment of the exchange rate and
thereby propitiate a rise in inflation.[39]
In all these classic approaches, an
excessive expansion of credit leads the public, national and/or foreign, into a
refusal to continue purchasing debt and in all of them a day of reckoning is
finally forced upon the government and society.
Bordo and Schwartz scroll the
experiences of currency crises dating from the XVIII century to Mexico’s recent
episode and find reassuring evidence to support the classical contention:
currency crises stem from inconsistencies between currency commitments and
internal prices, or impending wars.[40]
Literally “...the theory of self-fulfilling speculative attack may have
intellectual merit but contributed nothing to our understanding of real-world
events. In every crisis examined here, the fundamentals are more than adequate
to account for the actions of speculators.” [41]
This account brings us back into the
Mexican crisis. Which were its fundamental causes? All the factors listed above
made some contribution, but it will be argued that those truly essential or
that had the greatest significance were the combination of the exchange rate
regime [42] with a
rapid expansion of credit, a substantial part of which was of poor quality to
boot.[43] The
surge of bad credits is in turn explained by flimsy bank’s capitalization and
the failure to ensure that bankers met the “fit and proper” criteria to own or
to manage the institutions.
It would be incorrect to isolate a
factor, like the proportion of short-term government debt held by foreigners
whose holdings have been shown to be particularly volatile.[44] Such
volatility is probably derived from the ease with which, under a fixed or
quasi-fixed exchange rate, peso demand fluctuations have to be and are expected
to be readily accommodated.[45]
Volatility and risk stem in part from the exchange rate regime. A floating rate
presents speculators with currency uncertainty compounded by other risks,
notably market value risks.
The lower risk speculators confront
under a fixed exchange rate is borne by the government, i.e., by society at
large. The insurance premium paid by society to cover exchange rate risks is
proportional to the size of the international reserves needed to reassure
investors that potential claims will be satisfied. Mexico’s reserves were
insufficient even before December 1994 because of the size of the country’s
financial sector. Some authors puzzled by the depth and virulence of the
Mexican financial crisis have tried to explain it, at least partly, by pointing
out the financial vulnerabilities of the country. [46] In this
endeavor, Calvo compares Mexico’s public debt service requirements for 1995,
including amortizations, with those of Argentina, Brazil and Chile and relates
them to exports. The result is a rather high figure for Mexico, 160 per cent,
while the corresponding values for Argentina and Chile hovered around 50 per
cent. Brazil had a ratio similar to Mexico’s, but mostly non-volatile investors
held its debt: banks.
This line of reasoning is
insightful, but it does not go far enough. Although Calvo does relate M2 to
international reserves, the amounts involved at risk of a sudden demand shift
refer to a concept much wider than M2.[47] All
domestic and foreign liabilities, peso and foreign currency denominated, have
to be honored if there is a run on a country committed to a fixed exchange
rate. The amounts involved according to Mexico’s figures for 1994 are
staggering. At the end of November 1994, US$30 billion of short term bank debt,
25 billion of Tesobonos, 100 per cent of commercial banks liabilities to
resident claimants or 113 billion, the interest on all private and public
external debt due in 1995, approximately 11.7 billion, plus 14 billion of
currency in circulation.
Thus, total amount that all
potential claimants expected to collect under a run, at the prevailing exchange
rate, equaled roughly 200 billion dollars, or 45 per cent of the country’s GDP,
or 15 times the international reserves held at the end of November 1994. The
reason for adding up the concepts enumerated is that bank liabilities had full
government backing. Therefore, there is no justification to include only
Tesobonos, or these plus M2, nor to explain the exclusion of any
item belonging to bank liabilities nor any other government bond. One must also
take into account that about 70 per cent of all bank liabilities were payable
overnight and that the rest were very short term.
But this situation of extreme
liquidity was not new to Mexico or to most other countries. What was new was
the coexistence of a formidable growth in the volume and speed of international
capital movements [48] with
the persistence, in some countries, of a fixed exchange rate. In this regard
Mexico’s currency collapse was not much different from that of several European
countries in 1992.
Hence, the combination of a fixed
exchange rate, a substantial rise in US interest rates, a commitment to convert
to foreign exchange a substantial amount of financial assets and the political
tensions accumulated during 1994 comprise the true measure of Mexico’s financial
vulnerability at the end of 1994.
Most researchers of this period ignore that fractional reserve banking
requires a lender of last resort. Banks cannot liquidate loans when there is a
run. Because of this simple but inescapable fact, all the lines that have been
written about the so-called excessive expansion of the central bank’s internal
credit during 1994 are nonsensical. The logic of a fixed exchange rate is
implacable. When there is a run banks are, all of the sudden, left with more
loans than deposits. Hence, when the central bank lends to commercial banks to
balance their positions, it is simply fulfilling its unavoidable obligation as
lender of last resort. This chore is either performed by foreign creditors,
which is unlikely when there is a run, or by the central bank.
If commercial banks had had reserve
requirements or large liquidity coefficients the qualitative nature of the
argument remains. It matters little that in a run banks run down their reserves
at the central bank. In this latter case the accounting result is that the
central bank swaps dollars in exchange for the lower reserves held with it by
commercial banks, instead of providing the dollars to the commercial banks in
exchange for an increase of their liabilities at the central bank.
The fact that this latter operation
is labeled “central bank credit” and the other one is not is immaterial. In the
reserves at the central bank case,
the central bank’s liabilities decreased pari-passu
with its loss of international reserves, a result analogous to the so-called
increase in central bank credit. Finally, if [49]the
central bank requires banks to hold part of their liabilities in liquid foreign
assets, the central bank accounts may not even budge under a run, and yet the
qualitative and economic result is the same, the commercial banks lower
simultaneously an asset and a liability. In this case part of the nation’s
international reserves are held by commercial banks. That is the reason why,
even though there is a loss of hard-currency reserves, the accounts of the
central bank remain untouched.
To sum up the sequence of events
amply described and documented elsewhere,[50] the
chain reaction was initiated by an increase in the demand of dollars. This
increase had a counterpart in a lower demand for peso assets, and was followed,
as it turns out simultaneously because of the implacable mechanics of the daily
clearing of the payments system, by an increase in credit from the central bank
to the commercial banks.
Therefore, in circumstances such as
those described above, the increase in central bank credit is a fatal
consequence of the fall in the demand for pesos, it is a passive reaction, and
it is an unavoidable outcome of the lower demand for Mexican assets under a
fixed exchange rate regime.
References
1. Barro (1996) Monetary and Financial Policy. Getting
It Right. The MIT Press. Page 49.
2. Bordo and
Schwartz (1996). Why Clashes Between
Internal and External Stability Goals End in Currency Crises, 1797-1994. Washington
D.C.: NBER.
3. Borja, Gilberto (1996).CEO of Nacional
Financiera, quoted in El Economista, p.20.México D.F. (September 9).
4. Calvo, G.A., L.
Leiderman y C.M. Reinnhart (1993), Capital
Inflows and Real Exchange Appreciation in Latin America. IMF Staff Papers, 40.
5. Calvo,
Guillermo (1995) and Mendoza (1995) Reflections on Mexico’s Balance-of-Payments
Crisis. A Chronicle of a Death Foretold. Preeliminary
manuscript. College Park and Washington D.C.: University of Maryland and
Federal Reserve System (October).
6. Calvo,
Guillermo (1996), Capital Flows and
Macroeconomic Management: Tequila Lessons. International Journal of Finance
Economics. Vol. 1.
7. Diamond and
Dybvig (1983). Reprinted in Federal Reserve Bank of Minneapolis, 2000).
8. Fisher, Irving
(1930) The Theory of Interest
.Macmillan, New York.
9. Gil Díaz,
Francisco and Cartens, Agustín (1997). Pride and Prejudice: The Economics
Profession and Mexico’s Financial Crisis, 1994-1995
10. Gil Díaz,
Francisco and Cartens, Agustín (1996). Some Hypothesis Related to the Mexican
1994-95 Crisis. Mexico DF: Research Paper 9601, Banco de México (January).
11. Gil Díaz,
Francisco and Cartens, Agustín (1996) “One Year of Solitude: Some Pilgrim
Talles About the Mexican 1994-95 Crisis.” American Economic Review (June)
12. Gil Díaz,
Francisco (1995). A Comparison of Economic Crisis: Chile 1982, Mexico in 1995.
Paper presented at Forum 95 for Managed Futures and Derivatives, managed
Futures Association Chicago (July).
13. Ortiz-Martínez,
Guillermo (1994). La Reforma Financiera y
la Desincorporación Bancaria. México D.F.: Fondo de Cultura Económica.
14. Hale, David
(1995) Lessons from the Mexican Crisis of
1995 for International Economic Policy. Preliminary manuscript. Vienna:
Oesterreichische Nationalbank (September).
15. Hale, David
(1996) Lessons from the Mexican Crisis of
1995 for the Post Cold War International Order. The World Bank Report on
Mexico. Zurich-Kemper Investments Chicago.
16. Hausmann,
Ricardo and Eichengreen, Barry (1999) Exchange
Rates and Financial Fragility NBER Working Paper No. W7418.
17. Hausmann,
Ricardo and Gavin M. (1995). The Roots of
Banking Crises: The Macroeconomic Context”, paper presented at the
Conference on Banking Crises in Latin America, Washington D.C
18. Hayek
19. Honohan Patrick
(1997). Banking System Failures in
Developing and Transition Countries: Diagnosis and Prediction. Basle: BIS.
20. Honohan,
Patrick. 1996. Financial System Failures
in Developing and Transition Countries: Diagnosis and Prediction. Paper
prepared for the IMF/BIS/Basle Committee Conference “Strengthening the
Financial Systems in Developing Countries
21. Klingebiel,
Daniella. The Use of Asset Management in
the Resolution of Banking Crises, Cross-Country
Experiences. Working Paper. The World Bank. Wash., D.C. 2000.
22. Johnson, Harry
G. (1972) “The Monetary Approach to the
Balance of Payments” in the Journal of Financial and Quantitative Analysis
7.
23. Krugman, Paul
R.(1979) “A model of Balance of Payments
Crisis” in the Journal of Money, Credit and Banking 11,
24.
Krugman, Paul. (1998). “Saving Asia: It’s Time to get Radical”,
Fortune,
25. Krugman, Paul. (1999). Depression
Economics Returns. Foreign Affairs, 78, 1.
26. Lindgren (1996)
27. Lindgren,
García and Saal (1996),Bank Soundness and Macroeconomic Policy. Washington D.C.
: International Monetary Fund
28. Mancera, Miguel
(1997). Problems of Bank Soundness: Mexico’s Recent Experience. Seminar on
Banking Soundness and Monetary Policy in a World of Global Capital Markets.
IMF, January 28.
29. McKinnon,
30. McKinnon and
Pill (1995), Credible Liberalizations & International Capital Flows: The
Over borrowing Syndrome. Unpublished manuscript. Palo Alto, Calif.: Stanford
University.
31. Mundell, Robert
A. (1968). International Economics.
New York: Macmillan
32. Sánchez
Santiago, Emilio. Los Momios del IPAB,
El Economista Newspaper, 6/6/00
33.
Sargent Thomas. J and Neil Wallace (1981) “Some Unpleasant Monetarist Arithmetic”, Federal Reserve Bank of Minneapolis
Quarterly Review, (Fall).
34. Shelton, Judy
(1994). Money Meltdown:Restoring Order to the Global Currency System. New York:
The Free Press.
35.
Trigueros (1997) “Capital Inflows and Investment Performance: México.” Mexico City:
Centro de Análisis e Investigación Económica-Instituto Tecnológico Autónomo de
México.
36. Velasco, Andrés
(1987). “Financial Crises and Balance of
Payments Crises.” Journal of Development Economics 27.
37. Werner,
Alejandro. Un Estudio Estadístico sobre
el Compòrtamiento de la Cotización del Peso Mexicano frente al Dólar y de su
Volatilidad, Banco de México, Documento de Investigación 9701
38.
[1] The dollar, the Euro and the Yen are
internationally accepted currencies that can be issued on a last resort basis.
The IMF as a lender to “forestall or cope with an impairment of the
international monetary system”, GAB, or General Agreement to Borrow, feeds on
lending from the 11 large industrial countries. The IMF also has an issuer role
as a creator of supplementary reserves in the form of SDR (Special Drawing
Rights), although SDR issuance has been non-existent since 1970. (Bordo and
James, The International Monetary Fund: Its Present Role in Historical
Perspective, 2000). The moral hazard that arises from the role of international
institutions was not born out of the "Tequila” crisis as Bordo and James
suggest, witness the international, and IMF, support for France in 1968. (NBER,
Working Paper 7724).
[2] Intra-day
lending has been diminished or eliminated by payment systems that have evolved
towards the instantaneous settlement of every transaction. End-of-day lending
can also be reduced or eliminated, but with fractional reserve banking the
central bank will always lurk in the background as a potential savior under a
desperate liquidity need.
[3] This is a
misnomer. The total amount of credit remains constant even if the central
bank’s credit is negative, just as there is no credit injection when at the
other end of the exchange rate band the central bank ends up subtracting credit
as it tries to prevent a depreciation of the exchange rate. In this latter case
the central bank’s balance sheet would record an “expansion of internal credit”
with total credit, again, remaining constant.
[4] As much can be
said of the recent Mexican, the Russian, the Brazilian and the Southeast Asian
crises. The Scandinavian and UK crises of 1992 also fit the pattern.
[5] The same can
be said of the Asian crises: 1998.
[6] 1995 is not
included in the comparison because in that year the figures are contaminated by
the liquidation of Tesobonos by the Mexican government, the now infamous
dollar-linked peso securities first issued at the end of 1991.
[7]
Notwithstanding the political shocks, sooner or later the vulnerability derived
from enormous instantly callable liabilities was going to create a run. There
were several runs during the year that drained reserves to only 13 billion
dollars as of end of Nov. 1994. Then the dies were cast when at his Dec.
inauguration Zedillo shook confidence by his failure to confirm outgoing
Treasury Secretary Pedro Aspe (Barro,1996 and Bartley,2000) at the helm of the
Treasury: hard currency resumed its exit instead of flowing back in, until the
signal for the final attack was given by Zedillo’s personal decision to again
set a ceiling (ephemeral as it turned out) to the exchange rate, instead of
opting for a float and sealing some international reserves.
[10] In Mexico City real estate prices
increased 17.6 times in the December 1987 - December 1994 period, while the
consumer price index over the same time span increased 3.6 –fold. These asset
price bubbles coincided with a large expansion of mortgage credit into housing
and office building booms.
[11] There were
exceptions, such as Korea, which relied mostly on bank loans.
[12] Contrary to
some widely held perceptions; information concerning the behavior of the
Mexican economy was available to anyone who wanted to see it. Data on the balance of payments, the nature,
size, and volatility of capital flows, the size and speed of expansion of
credit, and the growth of the non performing portfolio of the banks were there
for anyone to see, with a timeliness and quality equaled even then by few
countries.
[14]
Gil Díaz (1999) Discussion Summary to
Edwards (1999) Capital Flows to Latin America. International Capital Flows.
NBER.
[15] The index of
the price of urban land in Mexico City (Table 1) shows its real value going
from 1,930.3 to 33,927.2 in the December 1987-December 1994 period, an increase
of 17.6 times in real terms.
[16] Hale (1995).
[17] Mancera
(1997).
[18] The demand for
peso assets has been shown to be highly sensitive to changes in the price of
the 30-year US Treasury bond that experienced a sharp rise in 1994. See Calvo,
G.A., L. Leiderman y C.M. Reinnhart (1993), “Capital Inflows and Real Exchange
Appreciation in Latin America”. IMF Staff
Papers, 40, p. 108-151.
[19] The banks’
last effort to cover their positions contributed to the fast depletion of
reserves prior to the float. The explanation lies in a foreign currency
liabilities-matching requirement that they had been allowed to satisfy in part
through holdings of Ajustabonos, securities
linked to the CPI. Understandably, the Ajustabono
position was largely eliminated in the week prior to the devaluation:
between December 15 and 23, the banks increased their dollar assets to the tune
of US$3.2 billion. The amount was substantial and originated in an earlier
speculation in Ajustabonos that had
led banks to hold inordinate amounts of them. Since banks could not issue
matching liabilities, allowing them to fund these assets with dollar
liabilities was an expedient way to prevent them from incurring losses.
[20] Honohan
Patrick (1997). Banking System Failures
in Developing and Transition Countries: Diagnosis and Prediction. Basle:
BIS. P.13.
[21] See Lindgren,
García and Saal (1996), p. 100. Also Honohan (1996): “Often hailed as the
panacea for banking weaknesses of one sort or another, privatization has all
too often been the regime change which incubated more serious problems. This
has been the case both in transition economies and in developing countries that
had operated with state owned banks. The problem has generally lain in the lack
of suitability or experience of the new owners, in the inadequate
capitalization of the privatized banks or both. (Honohan, Patrick. 1996. Financial System Failures in Developing and
Transition Countries: Diagnosis and Prediction. Paper prepared for the
IMF/BIS/Basle Committee Conference “Strengthening the Financial Systems in
Developing Countries.”)
[22] Ibid-p. 107.
[23] Ibid-p. 12.
[24] Ibid. P. 107.
[25] Ibid. P. 107.
[26] Ibid. P. 110.
[27] Chile’s
1982-83 crisis has many parallels with Mexico’s. See Gil-Díaz (1995) and
Velasco (1987) (Velasco, Andrés (1987). “Financial Crises and Balance of
Payments Crises.” Journal of Development
Economics 27, p. 263-283).
[28] Mancera (1997)
discusses the causes of the increase in private debt and provides a full
presentation of the diverse financial salvage operations required in the
aftermath of the crisis.
[29] “Unusual asset
price movements, rapid growth of lending, specially for property transactions
and for financing of stock market positions, capital inflows: these are some of
the tell-tale signals of a credit financed asset-price boom which may prove to
be unsustainable.” Honohan (1996), p. 13.
[30] Lindgren et.al. (1996), p. 77-78.
[31] Ibid. P. 84.
[32] All the
figures quoted in this section were provided directly by the Economic Research
Department of Banco de México.
[33] See Table
4.3.1.
[34] Borja,
Gilberto, Director of NAFINSA, the biggest development bank, for a brief period
that was interrupted when he made the statement quoted (1996).
[35] See Table 5.
[36] 4.74% / 5.83%.
[37] See Guillermo
Calvo (1995) Reflections on Mexico’s
Balance of Payments Crisis. A Chronicle of a Death Foretold.
[38] Bordo and
Schwartz (1996). Why Clashes Between
Internal and External Stability Goals End in Currency Crises, 1797-1994. Washington
D.C.: NBER.
[39] Johnson, Harry
G. (1972) “The Monetary Approach to the Balance of Payments” in the Journal of Financial and Quantitative
Analysis 7, p. 1555-1572. Sargent Thomas. J and Neil Wallace (1981) “Some
Unpleasant Monetarist Arithmetic”, Federal
Reserve Bank of Minneapolis Quarterly Review, (Fall). Krugman, Paul R., “A
model of Balance of Payments Crisis” in the Journal
of Money, Credit and Banking 11, (August). Mundell, Robert A. (1968). International Economics. New York:
Macmillan
[40] Bordo and
Schwartz (1996) p. 46.
[41] Honohan (1997)
p. 2 and 3, reaches similar conclusions.
[42] Bordo and
Schwartz (1996).
[43] Honohan (1996)
p. 13: “Unusual asset price movements, rapid growth of lending, specially for
property transactions and for financing of stock market positions, capital
inflows: these are some of the tell-tale signals of a credit financed
asset-price boom which may prove to be unsustainable.” Many other such quotes
and evidence are found in the literature. Rapid growth in bank lending is
considered “...a classic leading indicator of individual bank failure, and may
also serve for systems.” (Honohan. 1996. p.1 of Annex) and M. Gavin and R.
Hausman (1995) “The Roots of Banking Crises: The Macroeconomic Context”, paper
presented at the Conference on Banking Crises in Latin America, Washington D.C.
See also Honohan (1997), p. 3 to 6.
[44] See Guillermo
Calvo (1996), “Capital Flows and Macroeconomic Management: Tequila Lessons”. International Journal of Finance Economics.
Vol. 1, p. 120.
[45] Evidence of
how different institutional arrangements condition market behavior can be found
in the comparison of the 1988-89 adjustment period with the 1995-96 one (See
Alejandro Werner, Un Estudio Estadístico sobre el Comportamiento de la Cotización del Peso Mexicano frente al Dólar y de su Volatilidad, Banco de México, Documento de Investigación 9701). Werner found that the
volatility of interest rates and the average value of the real interest rate
were much lower in the latter period. Both intervals have several similarities,
the most important one being that both were phases of adjustment to a crisis,
but also a major difference: in the 1995-96 adjustment program a flexible
exchange rate was adopted versus a predetermined rate in the 1988-89 program.
[46] See Calvo
(1996) p. 208.
[47] One could even
counter argue that the lowest risk convertibility entails regarding the
possibility of a flight from the local currency is to be found in holdings
which may be totally liquid but that are needed for transactions (M0 or
M1) and, conversely, that the highest convertibility risk comes from
those bank and government obligations not included in M2.
[48] See Hale
(1996).
[49] The liquidity
coefficient.
[50] See Mancera
(1997), and Gil-Díaz and Carstens (1997).
No hay comentarios:
Publicar un comentario