Examining National Currency · Part III

Mexico: When the Peso Promise Broke

Four stages of a currency crisis—from capital inflows and political management to devaluation, rescue and recession.

The peso crisis did not begin on 20 December 1994, when Mexico widened its exchange-rate band. It had been accumulating behind an exchange rate designed to look calm. Foreign capital financed a widening current-account deficit. Reserves absorbed the first loss of confidence. Dollar-linked Tesobonos moved exchange-rate risk from investors to the Mexican state. When the promise finally broke, the price adjustment arrived everywhere at once.

A fixed or tightly managed price does not remove risk. It changes where risk waits—and who owns it when the price can no longer be defended.
Stage 1

The quiet regime

Capital inflows, credit and a large external deficit accumulate behind a controlled peso.

Stage 2

The defense

Political shocks drain reserves; domestic credit and Tesobonos postpone adjustment.

Stage 3

The break

The exchange-rate promise fails and market liquidity disappears.

Stage 4

The transfer

Rates, inflation, bank losses and recession distribute the cost through the economy.

The political and market timeline

NAFTA begins. The Zapatista uprising begins in Chiapas on the same day.

PRI presidential candidate Luis Donaldo Colosio is assassinated. Capital flight intensifies.

Ernesto Zedillo wins the presidential election. Reserve defense and short-term refinancing continue.

PRI secretary-general José Francisco Ruiz Massieu is assassinated.

Zedillo succeeds Carlos Salinas de Gortari.

The band is widened, confidence collapses, and the peso is allowed to float.

The United States announces a stabilization program centered on up to $20 billion from the Exchange Stabilization Fund.

Mexican Treasury-bill yields move above 69 percent; the stabilization cost reaches borrowers, banks and the real economy.

Stage 1 — Stability attracts capital

Mexico used the exchange rate as a nominal anchor. Privatization, financial liberalization and the prospect of NAFTA reinforced foreign confidence. Capital flowed in, credit expanded and imports grew faster than export receipts. By 1994 the current-account deficit was close to $30 billion—roughly 7 percent of GDP.

The arrangement could persist while foreign investors rolled their claims. Its apparent stability was therefore conditional. The band suppressed observed peso volatility, but it could not suppress the economic imbalance building underneath it.

The external deficit widens

$30bn$15bn$0 7.5242330 1990199219931994
Approximate current-account deficit, US$ billions; selected annual checkpoints. The deficit required continuing external finance.

Quiet prices can hide a jump process

7.2%94.2%19.5%7.1% Pre-breakBreak–MarRest 19951996
Annualized standard deviation of daily log changes in DEXMXUS. The low-volatility regime ended discontinuously.

Stage 2 — The adjustment is postponed

The uprising in Chiapas and the assassinations of Colosio and Ruiz Massieu changed how investors priced political stability. Banco de México sold dollars to defend the band. A fully unsterilized reserve loss would have contracted domestic money and forced rates higher. Instead, domestic credit replaced part of the reserves leaving the central bank.

The Treasury also shifted funding from peso-denominated Cetes toward Tesobonos, whose repayment was linked to the US dollar. Investors received protection against devaluation; the government assumed the currency risk. Outstanding Tesobonos approached $29 billion by the end of December 1994, while the liquid reserve cushion had fallen far below that amount. Mexico had bought time, but it had turned a currency problem into a sovereign rollover problem.

Reserves absorb the first run

$30bn$15bn$0 29.312.96.34.4 Feb 1994Nov 1994Dec 1994Jan 1995
Reserves excluding gold, approximate US$ billions; selected monthly checkpoints. Source series: FRED TRESEGMXM052N. The FRED page carries the IMF copyright and use notice.

Stage 3 — The promise breaks

On 19 December the peso closed near 3.47 per dollar. The next day, the incoming government widened the exchange-rate band. Rather than restore confidence, the action revealed that the previous promise could no longer be maintained. The peso closed near 3.95 on 20 December, reached about 4.85 when Mexico floated it on 22 December, and traded near 5.75 on 27 December. By 9 March 1995 it was around 7.60.

The repricing was measured in days

8.05.53.0 3.473.954.855.757.60 19 Dec20 Dec22 Dec27 Dec9 Mar
Mexican pesos per US dollar, selected observations. A higher number is a weaker peso. Source: FRED DEXMXUS.
From the options desk

In 1994 I was an OTC options market maker in Mexican ADRs. When the peso broke, the market did not wait for New York to open. I received calls in the middle of the night asking for prices.

The currency, local shares, ADRs, implied volatility and correlation were all moving while executable liquidity disappeared. That is what a credibility break looks like from the options desk.

Stage 4 — The cost moves through the economy

Once the exchange-rate anchor had gone, Mexico needed a new one. Short-term Treasury-bill yields rose from about 18.5 percent in December 1994 to 37.3 percent in January, 69.5 percent in March and 74.8 percent in April. Those rates helped stabilize the peso and attract capital back. They also increased defaults, damaged bank balance sheets and deepened the recession.

Treasury-bill yields

18.537.369.574.8 Dec 94Jan 95Mar 95Apr 95
Percent per annum; selected monthly checkpoints. Source series: FRED INTGSTMXM193N.

Real GDP growth

+4.9%−6.3%+6.8%+6.8% 1994199519961997
Approximate annual real GDP growth. The rebound did not erase the balance-sheet damage absorbed by households and banks.
Pressure pointBefore the breakAfter the breakWho absorbed the loss?
Exchange rate3.4662 pesos/$, 19 Dec 1994About 7.60, 9 Mar 1995Peso holders and unhedged borrowers
Daily FX volatilityAbout 7.2% annualizedAbout 94.2%, 20 Dec–31 MarDealers, ADR investors and options books
Treasury-bill yieldAbout 18.5%, Dec 1994About 74.8%, Apr 1995Borrowers, banks and taxpayers
Real GDP growth+4.9% in 1994−6.3% in 1995Workers, firms and households
Consumer pricesManaged under the peso anchorAverage inflation about 35% in 1995; year-end above 50%Household purchasing power

When the ruling machine captures the currency

It is tempting to compress the story into a criminal morality play. That goes too far. The monetary record does not prove that a crime syndicate literally controlled Banco de México, nor does it prove that Carlos Salinas expanded domestic credit with the specific intention of winning an election. Salinas himself was term-limited; the 1994 vote determined whether the PRI would retain the presidency under Zedillo.

But the interpretation cannot be dismissed. The state balance sheet was used to postpone adjustment through a politically critical year. Investors increasingly struggled to distinguish the credibility of the government, the ruling party and the presidential family. Raúl Salinas was arrested on 28 February 1995 in connection with the Ruiz Massieu murder—not for drug running. Separate investigations later examined the transfer of roughly $90–$100 million through Citibank affiliates and alleged money laundering. Allegations, arrests, convictions and later reversals should not be collapsed into one claim.

The deeper point is institutional. Once markets suspect that the monetary system is being managed around the political calendar, every reserve number, debt auction and official denial becomes evidence about the regime itself.

The modern echo

President Trump repeatedly demanded lower interest rates from Federal Reserve Chair Kevin Warsh. On 16 September 2026, the Warsh-led FOMC voted 12–0 to raise its target range by 25 basis points to 3.75–4.00 percent.

A president may demand cheaper money; independence exists only when the central bank can refuse. One meeting does not permanently establish independence. It demonstrates independence at the moment of decision.

Read the Federal Reserve statement →

Where the risk went

At each stage the policy did not eliminate the loss. It changed its owner.

StagePolicy or market mechanismRisk transferred to
1 · Capital inflowsThe stable peso finances credit, imports and an external deficit.A refinancing requirement dependent on foreign confidence.
2 · Reserve defenseDollar sales and domestic-credit expansion postpone contraction.The central-bank balance sheet.
2 · TesobonosDollar indexation protects investors from devaluation.The Mexican government and its taxpayers.
3 · DevaluationThe controlled price becomes a market price.Borrowers, banks, ADR holders and derivatives books.
4 · StabilizationVery high rates and external support restore a nominal anchor.Workers, firms, households and the public balance sheet.

Run the analysis

The RainbowStats script contains the complete four-stage slideshow, executable peso and CPI charts, the volatility study, timelines and tables.

Open the Mexico script in RainbowStats

View the underlying data on FRED

Some historical series supplied to FRED by the IMF or OECD cannot be redistributed through the public RainbowStats plug-in. The original observations, source notes and rights notices remain available on their FRED pages.

Readers should consult each series page for definitions, revisions, citation requirements and third-party terms of use.

Sources and interpretive boundaries

Figures are rounded. Current-account, reserve, yield and GDP graphics show selected historical checkpoints rather than reproducing restricted raw series. Claims about electoral motives and political capture are clearly identified as interpretation.

A currency is a political promise

Mexico suppressed the visible volatility of the peso until it could no longer suppress the underlying loss of confidence. The adjustment postponed through the election returned as devaluation, extreme interest rates and recession. Political pressure is not the same as political control. The test is whether the central bank can say no.