Latin American Debt Crisis: Mexico's 1982 Moratorium and the Lost Decade

On Friday 13 August 1982, Mexico's finance minister Jesús Silva Herzog boarded an Eastern Airlines flight from Mexico City to Washington with a single message for the US Treasury, the Federal Reserve, and the International Monetary Fund. His country, which had borrowed roughly $80 billion from commercial banks during the petrodollar decade, could not make its payment on Monday 23 August. By Monday, Silva Herzog had explained the same thing to roughly 800 commercial bankers crowded into the Federal Reserve Bank of New York's wood-panelled meeting room on Liberty Street. Within three months Brazil, Argentina, Chile, Venezuela, and another two dozen smaller debtors had stopped paying as well. About $330 billion of Latin American external debt — close to half of all developing-country borrowing on earth — entered a restructuring that would not be resolved until Nicholas Brady's 1989 plan finally let creditor banks accept a write-down.

The petrodollar engine

To understand August 1982, start with October 1973. The Yom Kippur War triggered an OPEC embargo, the price of Saudi Light quadrupled inside a quarter, and the cartel's annual current-account surplus jumped from $7 billion in 1973 to $68 billion in 1974. Those surpluses had to be parked somewhere. They flowed almost entirely into the eurodollar deposit base of the US and European money-centre banks — Citibank, Chase Manhattan, Manufacturers Hanover, Morgan Guaranty, BankAmerica, Lloyds, Deutsche, Crédit Lyonnais. Walter Wriston, who ran Citicorp from 1967 to 1984, famously argued that "countries don't go bankrupt", and he set his lending officers to redeploy the deposits as fast as they arrived. The geographical answer was obvious: oil-importing developing countries needed dollars to pay for the very oil that had filled the deposit base. Latin America, with growth rates running 5 to 7 percent annually and dictatorships hungry for infrastructure loans, became the natural counterparty.

For more on how the same oil shock reordered global finance, see the Yom Kippur War and the oil shock.

Between 1974 and 1981, syndicated lending to non-oil developing countries grew from about $14 billion outstanding to roughly $300 billion. Mexico's external debt rose from $19 billion in 1975 to $86 billion by mid-1982. Brazil's climbed from $24 billion to $87 billion over the same period. Venezuela's quadrupled. Almost none of the lending used fixed-rate paper. The terms were syndicated bank loans priced at LIBOR plus a margin — typically 0.625 to 1.5 percentage points over the three-month rate — with rollover dates every three or six months. Floating-rate exposure dominated the stock — 80 percent of Latin American medium- and long-term commercial debt in 1982 carried it (Cline, 1995). Every basis point of LIBOR move went straight onto the sovereign budget.

The Volcker shock arrives

In August 1979 Jimmy Carter installed Paul Volcker at the Federal Reserve to break double-digit US inflation. By October the new chairman had switched the Fed's operating procedure from targeting the funds rate to targeting non-borrowed reserves, letting the rate float wherever monetary aggregates demanded. The funds rate touched 20 percent in March 1980, fell back during a short 1980 recession, then climbed past 19 percent again in mid-1981. Three-month LIBOR moved in lock-step, rising from 8.8 percent in 1978 to 16.7 percent in 1981, and the dollar appreciated more than 50 percent on a trade-weighted basis between 1980 and early 1985. The mechanics are covered in the Volcker shock.

The arithmetic was vicious. Mexico's annual interest bill, which had been about $2 billion in 1977, hit $12 billion in 1982 on the same nominal stock of debt — an effect produced almost entirely by the LIBOR reset, not by new borrowing. Commodity prices, which had risen with the inflation of the late 1970s, fell sharply as US tightening pulled global demand down. The IMF commodity index lost a quarter of its value between 1980 and 1982. Oil — the asset that Mexican, Venezuelan, and Ecuadorian creditworthiness rested on — dropped from $35 a barrel in early 1981 to about $28 by mid-1982. The combined terms-of-trade and interest-rate shock between 1979 and 1982 cost Latin America roughly 15 percent of GDP annually in transferred resources (Sachs, 1989).

13 August, Washington

Silva Herzog spent Thursday 12 August in Mexico City calculating that the central bank had reserves of about $180 million left against a payment of roughly $300 million due on Monday. He phoned Treasury Secretary Donald Regan that night, then flew up Friday morning with Mexico's central bank governor Miguel Mancera. The Friday meetings ran from Treasury to the Federal Reserve to the IMF managing director's office. The IMF historian recorded Silva Herzog's opening line to Regan as a quiet statement that "we have run out of money" (Boughton, 2001). The package was assembled inside seventy-two hours: a $1 billion advance on US oil purchases for the Strategic Petroleum Reserve, a $1 billion Commodity Credit Corporation loan, $1.85 billion from the Bank for International Settlements through a swap arrangement with the New York Fed, and a commitment from Mexico to negotiate an IMF stand-by within thirty days.

Monday 23 August the IMF chaired a meeting of the thirteen largest commercial bank creditors — the so-called Bank Advisory Committee — at the New York Fed. Silva Herzog asked for a ninety-day rollover of all principal payments. The banks granted it, primarily because their alternative was to mark down syndicated loans that on average constituted between 80 and 200 percent of their primary equity capital. The exposure ratios were sobering (Cohen, 1991): Citibank's Latin American loans were 174 percent of equity at the end of 1982, Manufacturers Hanover's 263 percent, BankAmerica's 158 percent, Chase Manhattan's 169 percent. A simultaneous write-down to fair value would have rendered the largest US banks technically insolvent. Concerted lending, not honest accounting, was the only path the Fed and Treasury would tolerate.

The cascade

Brazil had been watching Mexico nervously since June. Brazilian banks rolled trade lines daily through the eurodollar interbank market in London and New York. By September 1982 interbank lines to Banco do Brasil's overseas branches stopped renewing. On 1 December, finance minister Ernane Galvêas flew to Washington for the same conversation Silva Herzog had begun in August. Argentina, fighting a war over the Falklands with the United Kingdom from April to June 1982, had already been frozen out of London markets and announced a sixty-day moratorium of its own in March 1982 — five months ahead of Mexico, though attracting less notice. Chile, whose finance officials had spent the late 1970s and early 1980s liberalising under Pinochet, watched its peso devaluation programme collapse in June 1982 as private dollar debt of its banks ballooned past 80 percent of GDP. Venezuela imposed exchange controls in February 1983, marking the end of its "saudi-style" oil-funded confidence.

Source: Federal Reserve Bank of New York and ECLAC, compiled by Cline, 1995.

The exposure column on the right matters more than the absolute debt totals. At the end of 1982, loans to the largest Latin American debtors were the single biggest credit concentration on the US banking system's balance sheet. William Seidman, who became FDIC chairman in 1985, later wrote that the regulatory community spent the years 1982 through 1988 in what he called "constructive concealment" — accepting bank accounting that valued non-performing sovereign loans at par while everyone privately understood the market price was thirty cents on the dollar.

The lost decade

What followed the rescheduling was not recovery. It was seven years of negative net resource transfers. Latin America, which had imported roughly $13 billion of foreign capital annually between 1973 and 1981, paid out an average of $24 billion a year to creditors between 1983 and 1989 — about 4 percent of regional GDP every year. Real per-capita GDP at the end of 1989 was lower than it had been in 1980 in Mexico, Argentina, Brazil, Venezuela, Peru, Bolivia, and most of the rest of the region. The Economic Commission for Latin America and the Caribbean coined the term "la década perdida" to describe what happened.

Mexico's experience was emblematic. Its inflation, which had run 28 percent in 1981, rose past 100 percent in 1982 and again in 1987. The peso was devalued from 26 to the dollar in February 1982 to 2,300 by 1988 — a factor of nearly ninety. Real wages in the formal sector fell by about 40 percent over the decade. The Mexican government cut public investment from 12 percent of GDP in 1981 to 4 percent by 1988 to free up dollars for debt service. Argentina suffered three hyperinflations — 1985, 1989, and 1990 — with monthly inflation in July 1989 hitting 197 percent. Peru defaulted on the IMF itself in 1985 under Alan García and went through a hyperinflation peaking at 7,650 percent for 1990. The political accompaniments were debt-fatigue elections that brought Argentina's Raúl Alfonsín, Brazil's Tancredo Neves, and Peru's García to power on platforms of restructuring or repudiation.

Baker fails, Brady delivers

Treasury Secretary James Baker tried in October 1985 — at the IMF annual meetings in Seoul — to convince the commercial banks to lend an additional $20 billion of net new money to the fifteen most-indebted countries over three years in return for IMF and World Bank structural reforms. The Baker Plan delivered roughly $13 billion of new lending against a stated $20 billion target, but produced no debt reduction. By 1988 the secondary-market price of Mexican loans had fallen to 40 cents on the dollar and Argentine loans to 22 cents. A formal paper on the debt-overhang problem showed analytically why net new lending could not solve a Pareto-dominated equilibrium (Krugman, 1988): if the existing stock could not be serviced in full, any rational marginal lender refused to add to it, and only a coordinated write-down could restore the borrower's incentive to invest.

Nicholas Brady, who had succeeded Baker at Treasury after the November 1988 election, delivered the speech on 10 March 1989 at a Bretton Woods commemoration conference. Brady proposed that commercial banks would exchange their existing loans for either a 30-year par bond carrying a below-market 6.25 percent coupon, or a 30-year discount bond priced at 65 cents on the dollar but paying a floating LIBOR-plus spread. Both bond types had their principal collateralised by zero-coupon US Treasury strips purchased with IMF, World Bank, and Japanese Eximbank money. Banks could pick the menu item that suited their tax and capital position.

Mexico signed the prototype Brady agreement on 4 February 1990. Roughly $48 billion of syndicated bank claims were exchanged: about 41 percent of creditors chose discount bonds, 49 percent chose par bonds with the reduced coupon, and 10 percent chose to provide new money instead. Present-value debt reduction worked out to about 35 percent of the restructured portfolio (Cline, 1995). Mexican equity prices rose 26 percent in the week after the deal closed. The peso stabilised; inflation came down from 30 percent in 1990 to under 10 percent by 1993. Capital began to return.

By 1997 eighteen countries had executed Brady exchanges, covering Argentina, Brazil, Bulgaria, Costa Rica, Côte d'Ivoire, the Dominican Republic, Ecuador, Jordan, Mexico, Nigeria, Panama, Peru, the Philippines, Poland, Russia, Uruguay, Venezuela, and Vietnam. The aggregate face value of Brady bonds issued exceeded $160 billion. Trading volumes in the secondary market grew from negligible in 1989 to about $2.7 trillion in 1996. The Brady issue catalogued the building blocks of every emerging-market sovereign bond that followed: standardised covenants, collateral, and listing on the Luxembourg and London exchanges. The 1994 Mexican peso crisis, which is covered in the 1994–1995 Tequila crisis, and the Argentine collapse of 2001–2002, played out inside that market structure.

What the loans bought, and what they cost

Looking back from 1990, the petrodollar lending boom of 1974 to 1981 had purchased almost nothing of lasting Latin American capital formation. The resource transfer into the region during the boom years had largely been offset by capital flight (Sachs, 1989) — Mexican and Venezuelan residents had moved an estimated $60 billion out of their own banks between 1979 and 1982, often to dollar accounts at the same money-centre banks that were simultaneously lending to their governments. The international banking system had effectively recycled the petrodollars through Latin American sovereigns and back into the deposit base of the lenders, charging a 100 to 150 basis point spread on the round trip. The lost decade was the price of getting the round-trip unwound.

Silva Herzog, asked in a 1992 oral history what he remembered most clearly about August 1982, said that when he reached the Federal Reserve Bank of New York on Friday morning, Anthony Solomon, then president of the New York Fed, met him at the elevator and offered coffee. Then Solomon told him the institution had been waiting six months for him to arrive.