In January 2002, Argentina’s public debt jumped from roughly 62% of GDP to approximately 164% \u2014 not because the government had borrowed more money, but because the peso stopped being worth a dollar. The Convertibility Plan, which had pegged the two currencies 1:1 since 1991, collapsed. The peso fell to roughly 4 pesos per dollar. The external debt, denominated in dollars, remained the same number of dollars. But in peso terms, that same debt was now four times heavier. No new borrowing. No new spending. Just the exchange rate moving, and a country suddenly crushed under obligations that had grown almost three times over in the currency its people actually earned, saved, and paid taxes in.<\/p>\n\n\n\n
The arithmetic is simple. The mechanism is not accidental.<\/p>\n\n\n\n
Start with an abstraction. A country owes $10 billion to foreign creditors. The exchange rate is 2 local units to 1 dollar, so the debt equals 20 billion in local currency. The government depreciates its currency, or the market forces depreciation, or the peg breaks \u2014 and the exchange rate moves to 4:1. The same $10 billion is now 40 billion local units. The debt doubled, in the currency that matters domestically, with no new borrowing, no new deficit, no new political decision. Pure arithmetic operating through a structural condition.<\/p>\n\n\n\n
Apply this to Argentina. By late 2001, Argentina held approximately $93 billion in external dollar-denominated debt \u2014 at the time the largest sovereign default in history. When the peso broke from its 1:1 peg and fell to approximately 4:1, the local-currency weight of that obligation tripled. GDP shrank 11% in 2002. Unemployment exceeded 20%. Real wages fell 18%. More than half the population fell into poverty \u2014 not over years, but in months. The 62%-to-164% debt ratio jump summarizes a social catastrophe in a single data point.<\/p>\n\n\n\n
The mechanism doesn’t only activate in crises. It operates continuously, in both directions. When a country’s currency weakens modestly against the dollar \u2014 say 10% over a year \u2014 the local-currency value of its dollar debt increases by the same percentage. When the Federal Reserve raises interest rates, capital tends to flow toward dollar assets, strengthening the dollar against developing-economy currencies, and simultaneously increasing the local-currency burden of dollar-denominated debt across dozens of countries at once.<\/p>\n\n\n\n
The Fed’s March 2022\u2013July 2023 rate-hiking cycle raised the federal funds rate by 525 basis points. This was US domestic monetary policy, aimed at US inflation, shaped entirely by US economic conditions. But as the dollar strengthened, the Turkish lira, the Argentine peso, the Colombian peso, and the Egyptian pound each fell 15% or more against the dollar in 2022 alone. Every percentage point of dollar strengthening increased the local-currency debt burden in each of those countries \u2014 automatically, with no new borrowing, through the same arithmetic that operated in Buenos Aires in January 2002.<\/p>\n\n\n\n
This rarely appears in coverage of developing-country debt crises as a proximate cause. The standard frame attributes crises to fiscal mismanagement, governance failures, or policy errors. Those factors are sometimes real. But the amplifier was running before the crisis, will run after the crisis, and will keep running regardless of what any government does \u2014 because it’s built into the architecture of who gets to borrow in what currency.<\/p>\n\n\n\n
Why do developing countries borrow in foreign currencies at all? The obvious explanation is governance: lenders won’t trust currencies from countries with poor inflation records. It implies the condition is earned and therefore curable \u2014 fix your inflation record, rebuild your credibility, and lenders will eventually accept your currency. It locates the problem with the borrowers.<\/p>\n\n\n\n
In 1999, Barry Eichengreen and Ricardo Hausmann gave this condition a name: “original sin” \u2014 the situation where a country’s domestic currency “cannot be used to borrow abroad or to borrow long-term even domestically.” A country with original sin cannot issue international debt in its own currency. It has to borrow in someone else’s, which means the exchange rate amplifier is always running.<\/p>\n\n\n\n
In their 2002 paper with Ugo Panizza, “Original Sin: The Pain, the Mystery, and the Road to Redemption,” Eichengreen, Hausmann, and Panizza tested the governance explanation against the data. They found it didn’t hold. Inflation history was not a statistically robust predictor of whether a country could borrow internationally in its own currency. Countries with decades of price stability couldn’t escape original sin. Countries that had done everything “right” by the conventional metrics were still borrowing in dollars.<\/p>\n\n\n\n
The one variable that robustly predicted original sin was country size \u2014 specifically, the size of the economy relative to global financial markets. Large economies could issue internationally in their own currencies. Small and medium economies mostly couldn’t. In the 1993\u20131998 period, developing countries accounted for approximately 10% of international debt but less than 1% of its currency denomination. That is: developing countries issued international debt, but almost none of it was in their own currencies.<\/p>\n\n\n\n
The five exceptions to original sin \u2014 the countries that could borrow internationally in their own currencies \u2014 were the United States, the euro area (represented in the 1993\u20131998 data primarily by Germany and France, the euro having launched in 1999), Japan, the United Kingdom, and Switzerland. Not coincidentally, these are precisely the economies whose currencies serve as global reserve currencies or major trading currencies. Their advantage isn’t that they behaved better. It’s that they’re large enough that international investors have reason to hold their currencies directly, creating genuine demand for assets denominated in those currencies.<\/p>\n\n\n\n
Original sin is not a punishment for bad behavior. It’s an entry condition determined by economic mass. It was true of countries before they had bad inflation records. It stays true after they clean them up. The mechanism through which exchange rate movements amplify debt is not a consequence of governance failure \u2014 it’s the operating environment that governance failures occur within.<\/p>\n\n\n\n
On original sin redux<\/strong>\n\nSince the mid-2000s, some emerging markets have issued bonds denominated in their own currencies to international investors \u2014 a development sometimes read as original sin dissolving under the pressure of financial innovation. Agust\u00edn Carstens and Hyun Song Shin wrote in Foreign Affairs in March 2019 that this progress was real but incomplete. They coined \"original sin redux\": the currency mismatch hadn't disappeared, it had migrated. When a country issues local-currency bonds to foreign investors, those investors hold the currency risk. And when the local currency falls, foreign investors sell, generating capital outflows that further pressure the currency \u2014 a self-reinforcing exit dynamic that reproduces the original problem through a different channel. Boris Hofmann, Nikhil Patel, and Steve Pak Yeung Wu formalized this in BIS Working Paper No. 1004 (February 2022), building a model of how local-currency issuance had peaked between roughly 2008 and 2011 and partially reversed since. The currency mismatch never went away. It found a new address.<\/code><\/pre>\n\n\n\nArgentina: the repeating trap<\/h3>\n\n\n\n
The Convertibility Plan of 1991 was not a reckless gamble. It was a reasonable response to a real crisis. Argentina had suffered hyperinflation that reached approximately 3,000% annually in 1989. The dollar peg ended the hyperinflation. That was its achievement, and the achievement was genuine. But the peg locked Argentina into accumulating dollar-denominated external debt throughout the 1990s, because it had no choice \u2014 borrowing internationally meant borrowing in dollars, peso receipts couldn’t be used to issue dollar-equivalent bonds without the peg providing nominal certainty, and the peg required maintaining dollar confidence by borrowing to cover current account deficits. By the end of the decade, the architecture had become a trap.<\/p>\n\n\n\n
By 2001, external debt stood at approximately $93 billion. The IMF was pressing for fiscal adjustment \u2014 deficit reduction, public sector wage cuts \u2014 to maintain the conditions under which the peg could hold. The fiscal adjustment deepened the recession. The deepening recession reduced tax revenues and widened the deficit, requiring more adjustment. The cycle closed on itself. In late 2001, the IMF refused to release a scheduled tranche, accelerating the crisis. The peg broke in January 2002. The peso reached 4:1 against the dollar within months. Public debt went from 62% to 164% of GDP in the time it took an exchange rate to move.<\/p>\n\n\n\n
That’s 2001. Now 2018.<\/p>\n\n\n\n
By 2018, Argentina had made genuine fiscal adjustments. The policy environment was meaningfully different from the chaos of the 1990s \u2014 worth noting precisely because the conventional story needs a villain and fiscal irresponsibility is the standard candidate. But Argentina’s peso still lost more than 50% of its value in the first eight months of 2018. The government secured a $50 billion IMF loan \u2014 the largest in IMF history at the time, later expanded to $57 billion. The crisis happened anyway. The Eurodad network (European Network on Debt and Development), analyzing the program’s conditions, found the 2018 arrangement contained more conditions than the 2000 program \u2014 this despite IMF claims of having internalized lessons from the previous crisis. The trap reasserted itself not because governance failed more egregiously in 2018 than in 1990, but because the structural currency mismatch had not been addressed and cannot be addressed by national policy reform alone.<\/p>\n\n\n\n
Argentina's serial defaults<\/strong>\n\nArgentina has defaulted on its sovereign debt nine times. The standard reading is that this record reveals serial irresponsibility \u2014 a political culture incapable of maintaining debt sustainability. The structurally accurate reading is that Argentina keeps reencountering a trap that its own reforms cannot remove. The currency mismatch means that any external shock sufficiently large to weaken the peso will automatically enlarge the dollar-denominated debt in local-currency terms, potentially beyond the government's fiscal capacity regardless of how responsibly it has been managing its finances. Repeated IMF recourse is not primarily evidence of failed reform. It's evidence of a structural condition that keeps recreating the conditions for crisis.<\/code><\/pre>\n\n\n\nTurkey: same trap, different architecture<\/h3>\n\n\n\n
The Turkish crisis of 2018 illustrates something the Argentine case can obscure: original sin doesn’t require a sovereign debt crisis to express itself. By 2018, Turkey’s problem wasn’t primarily the government’s balance sheet. It was the corporate sector’s.<\/p>\n\n\n\n
Turkish companies had accumulated approximately $181 billion in short-term foreign currency debt by 2018. Construction \u2014 a sector with revenue almost entirely in Turkish lira \u2014 held roughly 70% of its obligations in foreign currencies; manufacturing was close to half.<\/p>\n\n\n\n
The companies earned lira and owed dollars. Same mechanism; different balance sheet.<\/p>\n\n\n\n
In August 2018, the lira lost 33.7% of its value against the dollar in a single month. The banks immediately faced pressure from corporations that couldn’t service their foreign currency obligations on lira revenues. Restructuring demands flooded in. What looked like a banking-sector problem was the exchange rate amplifier activating at the corporate level rather than the sovereign level.<\/p>\n\n\n\n
Then came 2021. President Erdogan dismissed central bank governor Naci A\u011fbal in March \u2014 A\u011fbal had been raising rates to fight inflation \u2014 replacing him with \u015eahap Kavc\u0131o\u011flu, who promptly cut the policy rate from 19% to 14% over several months. The lira fell immediately on the dismissal announcement and continued falling. By the end of 2021, it had lost 44% of its value for the year. December inflation hit 36%.<\/p>\n\n\n\n
Erdogan’s monetary theory \u2014 that high interest rates cause inflation rather than suppress it \u2014 is wrong. Removing central bank independence in a country with pre-existing dollar exposure was reckless. The decision to fire A\u011fbal was a serious governance failure, and calling it anything else would be false precision. But the analysis cannot stop there, because the damage was catastrophically disproportionate to the policy error precisely because the corporate sector was already saturated with foreign currency debt. The same interest rate decision in a country that could borrow internationally in its own currency would have produced inflation, currency weakness, and possibly a recession. It would not have produced an economy-wide balance sheet crisis in which every point of lira depreciation automatically increased the real burden of hundreds of billions in corporate obligations.<\/p>\n\n\n\n
Governance failure explains when and how the 2021\u20132022 crisis unfolded. Structural currency mismatch explains why the damage was as total as it was. These are different claims, working at different levels. Using one to exclude the other is not analysis \u2014 it’s selection bias in favor of the explanation that assigns all blame to the government.<\/p>\n\n\n\n
Which leaves the institution designed to catch these crises \u2014 and the question of what it actually does with them.<\/p>\n\n\n\n
The IMF and the architecture of the trap<\/h3>\n\n\n\n
The IMF’s crisis-resolution logic isn’t irrational on its own terms. Lend at below-market rates in exchange for fiscal adjustment, which is supposed to restore investor confidence, which is supposed to restore access to international capital markets, which is supposed to allow the country to roll over its debt and resume growth. Each step follows from the last. The structural problem is that the exit from crisis runs directly through the dollar system. Restoring investor confidence means restoring access to dollar-denominated international capital markets. The country emerges from crisis re-exposed to the same mechanism that produced the crisis.<\/p>\n\n\n\n
In Argentina in 2001, IMF conditionality required deficit reduction and commitment to the peg \u2014 deepening the recession while the peg was under lethal pressure from a current account deficit the peso’s overvaluation was generating. In Argentina in 2018, the $50\u201357 billion program \u2014 the largest in IMF history \u2014 attached more conditions than the program that preceded the 2001 collapse, by Eurodad’s analysis. The conditionality isn’t evidence of institutional malice. The IMF is not acting against Argentina’s interests as it understands them. It’s an institution whose entire logic of crisis resolution is built around restoring countries’ access to dollar-denominated international capital. That’s the toolkit. Within that toolkit, re-entering dollar debt markets isn’t a failure mode \u2014 it’s the intended outcome. The problem is the toolkit, not the people using it.<\/p>\n\n\n\n
Zambia's 2020 default<\/strong>\n\nIn November 2020, Zambia became the first African sovereign to default during the COVID era, missing a $42.5 million Eurobond coupon payment. The kwacha had depreciated approximately 64% against the dollar in 2020 \u2014 from roughly ZMW 12.9 to ZMW 21.1 per dollar. Zambia's external debt consisted largely of Eurobonds \u2014 denominated in dollars and euros \u2014 along with Chinese loans contracted in foreign currencies, with Eurobond issuances from 2012 to 2015 totaling $3 billion and Chinese supplier credits exceeding $4 billion by 2021. As with Argentina and Turkey, the exchange rate amplifier ran continuously: every point of kwacha depreciation increased the local-currency weight of dollar obligations. Debt relief negotiations stretched over years, complicated in part by the absence of any comprehensive international framework for sovereign debt resolution that could accommodate the mix of creditors. Zambia finally reached a restructuring agreement in 2023, after three years. The drawn-out process reflected not fiscal irresponsibility but the structural gap between where debt distress happens \u2014 developing economies with foreign-currency liabilities \u2014 and where the institutions for resolving it were built, if they were built at all.<\/code><\/pre>\n\n\n\nWhat reform actually requires<\/h3>\n\n\n\n
If the problem is the toolkit, what would a different one look like? The partial escape route \u2014 local-currency bond issuance \u2014 has already been shown to relocate the problem rather than solve it. The currency mismatch migrates to foreign investors’ balance sheets, and when those investors exit under stress, capital outflows drive further depreciation.<\/p>\n\n\n\n
Eichengreen and Hausmann proposed currency basket indices that would allow international investors to take diversified exposure to a portfolio of emerging market currencies without making single-currency bets. This would potentially create genuine demand for EM-currency assets without concentrating risk on any single balance sheet. The proposal hasn’t been adopted at scale. It requires coordination among institutions that don’t have strong incentives to coordinate, and it would reduce the demand for dollar-denominated assets, which the US benefits from.<\/p>\n\n\n\n
The more fundamental reform was proposed in 1944 and rejected before it could be implemented.<\/p>\n\n\n\n
At the Bretton Woods conference, John Maynard Keynes proposed a supranational currency \u2014 the bancor \u2014 for international trade settlement, to be issued by an International Clearing Union. No single nation’s monetary policy would generate reserve currency privilege, because the reserve currency would be issued by an institution accountable to all trading nations rather than one of them. The system would impose symmetric adjustment pressure: countries running persistent trade surpluses would pay a “bancor tax,” penalizing creditor nations as well as debtor nations for imbalances. The adjustment burden would fall on both sides of a trade relationship. Under Keynes’s proposal, the United States’ persistent trade surplus in the early postwar period would have generated automatic pressure on the US to reduce it \u2014 rather than allowing the US to run balance-of-payments deficits while foreign central banks accumulated the resulting dollar reserves, structuring the global financial system around a currency it alone controlled.<\/p>\n\n\n\n
The US rejected the Keynes plan at Bretton Woods. Harry Dexter White’s counterproposal \u2014 which anchored the entire system to the dollar \u2014 prevailed. This wasn’t a neutral technical choice about which clearing mechanism was more efficient. It was a choice about which country would extract what Val\u00e9ry Giscard d’Estaing, then Finance Minister of France, described in 1965 as the “exorbitant privilege” of reserve currency status: lower borrowing costs, the capacity to run deficits that would trigger debt crises in any developing economy, the ability to conduct monetary policy for domestic purposes regardless of the consequences for those holding dollar-denominated obligations globally.<\/p>\n\n\n\n
The US extracted that privilege in 1944. It extracts it now, every time the Federal Reserve adjusts interest rates and the dollar moves and the local-currency burden of debt in dozens of developing economies adjusts with it. Genuine reform \u2014 a multipolar reserve currency system, or a Keynes-style clearing institution imposing symmetric adjustment on surplus and deficit countries alike \u2014 would require the US to accept meaningful reduction in that privilege. Nothing in the historical record suggests it’s willing to. The countries with the most to gain from reform have the least power to achieve it. The countries with the power to block reform are those currently extracting the benefit. This is not a failure of international institutions. It’s a predictable outcome of how power is distributed within them.<\/p>\n\n\n\n
The 1997 Asian financial crisis<\/strong>\n\nThailand's baht had been pegged to the dollar. By mid-1997 the peg was under speculative attack, sustained in part by dollar-denominated corporate debt accumulated during the investment boom of the preceding decade. Thailand unpegged the baht in July 1997. The baht fell sharply. Dollar-denominated obligations, expressed in baht, became catastrophically expensive. Indonesia, South Korea, and Malaysia were drawn into the same dynamic within months \u2014 each had accumulated foreign-currency corporate debt during the boom period, each suffered the same exchange rate amplification of that debt burden when their currencies fell under contagion pressure. The IMF's response \u2014 austerity measures, sharp interest rate increases, structural adjustment conditions \u2014 was contested at the time and has remained contested. Joseph Stiglitz, who served as chief economist at the World Bank during the crisis, argued in \"Globalization and Its Discontents\" (2002) that IMF-imposed fiscal retrenchment made serious recessions deeper, and that interest rate increases generated cascading bankruptcies that worsened rather than restored confidence. The debate about whether IMF policy deepened or merely failed to prevent the contraction continues among economists. What isn't debated is that the mechanism producing the initial debt amplification \u2014 currencies falling against dollar-denominated obligations \u2014 was the same mechanism operating in Argentina and Turkey, because it's always the same mechanism.<\/code><\/pre>\n\n\n\nThe conventional narrative and its beneficiaries<\/h3>\n\n\n\n
Return to the arithmetic of early 2002. The peso had been worth a dollar. It was now worth a quarter of a dollar. The public debt had been 62% of GDP. It was now 164% of GDP. GDP had fallen 11%. More than half the population had slipped into poverty.<\/p>\n\n\n\n
Hold two facts simultaneously: Argentina’s fiscal management through the 1990s was often chaotic, politically constrained, and sometimes reckless. And the 2002 debt crisis was primarily generated by a mechanism entirely outside any Argentine government’s control \u2014 an exchange rate amplifier built into a currency mismatch that Argentina could not escape regardless of how it managed its finances, because original sin is determined by economic size and the structure of international financial markets, not by the quality of domestic governance.<\/p>\n\n\n\n
The conventional narrative keeps only the first fact. It reads the crisis as a consequence of Argentine failure, recommends austerity and conditionality as the cure, and produces a fresh round of dollar borrowing as the exit from crisis \u2014 which re-exposes the country to the same mechanism in the next cycle. This is not a narrative error. It’s a narrative that serves certain interests very well. It locates the cause of debt crises in the countries that suffer them, insulating the structural features of the international monetary system from examination. It justifies conditionality as a rational response to governance failure, maintaining the logic of an institution whose toolkit is built around re-integrating countries into the dollar system. It allows reserve currency holders to continue extracting exorbitant privilege without that extraction appearing in any standard account of what produced the crisis.<\/p>\n\n\n\n
Which countries benefit from the story being told the way it’s told?<\/p>\n\n\n\n
The architecture that tells it was built in 1944. The answer has been embedded there ever since.<\/p>\n\n\n\n