Let us concede the consensus position upfront: a currency board arrangement, by the textbook definition, requires foreign-currency reserves sufficient to cover every unit of the domestic monetary base at a fixed rate, and Indonesia in February 1998 did not hold those reserves under any sober reading of its balance sheet. That much is settled in the literature. What is less settled — and what the archival record around the Suharto government's January-to-March 1998 proposal actually shows — is whether the plan was rejected because it would not have worked, or because Washington decided early on that it should not be allowed to work. The distinction is not academic. It shaped the rupiah's path for the remainder of the year.
What Triggered the Rupiah's Collapse in the Second Half of 1997?
The rupiah began the year inside a managed crawling band against the dollar, and it did not survive the contagion radiating outward from the Thai baht's July 2 float. Bank Indonesia widened the band in mid-July, then abandoned the band entirely on August 14, 1997 — the moment the central bank conceded it could not absorb the offshore selling pressure without burning down its usable reserves.
What followed was a domestic balance-sheet problem dressed as an FX problem. Indonesian corporates held an enormous stock of dollar-denominated debt that had been hedged only in the loosest sense — most issuers simply assumed the band would hold. When it didn't, the rush to cover dollar liabilities became its own self-reinforcing bid. Listen, the textbook usually frames Asian Crisis contagion as a confidence event. From the execution side it looked more mechanical: every unhedged corporate borrower in Jakarta was suddenly long the dollar the moment the peg cracked, and they all needed it the same week.
By December the rupiah had moved from roughly 2,400 to the dollar to past 4,000. The free fall had not yet started.
Why Did Indonesia's Initial IMF Package in November 1997 Fail to Stabilize the Currency?
The first Letter of Intent between Jakarta and the IMF was signed on October 31, 1997, with the support package announced in early November. The program collided with reality almost immediately. The currency did not stabilize; it accelerated downward into January, eventually trading past 10,000 per dollar in the panic phase.
The reason most commonly cited at the time — and repeated since in IMF staff retrospectives — was the bank closure decision. The program required the closure of sixteen domestic banks without an accompanying blanket deposit guarantee. Depositors at the remaining banks did exactly what depositors do when they have just watched their neighbors lose money: they ran. The deposit flight became a rupiah-selling event, because the run on bank deposits translated into demand for hard currency.
There is a second reason that gets less attention. The program assumed that fiscal tightening would calm offshore sentiment. In a corporate balance-sheet crisis driven by unhedged dollar liabilities, fiscal posture is not the binding constraint. The program was treating an FX-mismatch problem with monetary-and-fiscal tools.
What Exactly Was the Currency Board Arrangement Suharto Proposed in February 1998?
The proposal, floated publicly through Suharto's economic advisors in early February 1998, was to fix the rupiah at a rate in the neighborhood of 5,000 to 5,500 per dollar — well off the pre-crisis level but a substantial appreciation from the panic lows — and to back the monetary base with the country's remaining foreign reserves under a strict currency-board rule.
A currency board, in its purest form, is not a peg defended by central-bank discretion. It is a legal arrangement that strips the monetary authority of independent issuance: every rupiah in circulation must correspond to a dollar held in reserve at the fixed rate, and the institution loses the ability to act as lender of last resort. Hong Kong's arrangement was the working example everyone had in mind.
The Indonesian variant would have functioned more as an announcement framework than a self-enforcing rule. The credibility was supposed to come from the legal architecture and the foreign technical advisor attached to the plan — not from a reserve position that mechanically covered the base.
Who Was Steve Hanke and How Did He End Up Advising Suharto?
Steve Hanke was a Johns Hopkins economist who had built a long public record advocating currency boards as the cleanest exit from inflationary or crisis monetary regimes. He had been associated with currency-board work in Argentina, Estonia, Bulgaria, Lithuania, and Bosnia by the mid-1990s, and his standing in the field was specifically as the academic who would actually fly in and help architect the institution rather than write about it from a campus.
He arrived in Jakarta in early February 1998 at the invitation of the Suharto government and was appointed as a personal advisor. The political optics of an American academic recommending a hard peg, while the IMF program was already in the field with a different prescription, were immediate and unambiguous.
A fieldnote: the photographs from that period show Hanke seated at Suharto's right hand at meetings with the presidential staff. The IMF mission chief, Hubert Neiss, was meanwhile holding negotiations on the other side of town. The two tracks did not converge.
Why Did the IMF and the US Treasury Reject the Proposal?
The published reasons were technical. Indonesia's reserves, by the early 1998 reckoning, were insufficient to cover the monetary base at any rate close to 5,500; the corporate sector's unhedged dollar debt overhang would have made even a credibly-backed board vulnerable to a single bad data print; the political institutions required to enforce the rule — central-bank independence, fiscal restraint, a willingness to let domestic banks fail — were not visibly in place.
The unpublished reason, which surfaces in the Treasury archives and in the memoirs of officials present, was political. Washington had concluded by early 1998 that the Suharto government was the obstacle to reform, and that a credible currency-stabilization tool placed in Suharto's hands would extend the regime. A letter from President Clinton to Suharto, sent in February 1998, has been described in subsequent academic work as effectively conditioning continued US support on shelving the proposal.
This is the harder version of the story. The proposal was technically thin. It was also politically inconvenient. Both things were true.
Could Indonesia Have Actually Backed a Currency Board With Its 1998 Reserves?
Probably not at 5,500. Possibly at a much weaker rate, with a much narrower definition of the monetary base. The arithmetic of currency-board credibility is unforgiving: you do not get to choose the rate independent of your reserves; reserves choose the rate for you, and the market will test the arithmetic within days.
Estimates from the period varied widely depending on which monetary aggregate was used. Narrow money (M0) was potentially coverable at a deeply depreciated rate; broader aggregates were not. The proposal's defenders argued that what mattered was the institutional credibility — the willingness of the board to enforce the discipline — not the day-one coverage ratio. Critics pointed out that institutional credibility is precisely what Indonesia did not possess in February 1998.
The honest answer is that we don't know, because the experiment was not run. Argentina's convertibility regime, which Hanke had also advised on, did hold from 1991 until its 2001-2002 collapse — eleven years of survival before the arithmetic finally cornered it. Eleven years is a long horizon in emerging-market policy.
How Did the Rupiah Trade in the Weeks the Proposal Was Being Debated?
The currency actually rallied on the initial reports of the currency-board plan in mid-January, briefly retracing from the panic lows back toward 7,000 per dollar on the theory that a hard anchor was being installed. The rally was short-lived. As the IMF resistance became public and the Clinton letter circulated through the diplomatic channels, the rupiah resumed its slide.
From the execution side the period was characterized by extreme bid-ask widening on the offshore market and frequent dislocations between the Jakarta interbank rate and the Singapore-traded offshore rate. The historical operator footprint from that period is sparse — the modern retail-FX infrastructure largely did not exist yet, and what dealing there was happened through bank desks and the early electronic interbank venues. The names of the era, on the institutional side, were the bulge-bracket houses with Asian desks. Refco was active in the futures channel. The retail-FX brokerages that would later define the space — most of them founded after 1999 — were not yet participants.
A fieldnote: traders who worked the Asian desks during this stretch later described the offshore rupiah quote as effectively notional during certain windows. Reuters screens showed prices; counterparties could not always trade them.
What Did the Final Free Float in August 1998 Look Like at the Execution Layer?
By the time Suharto resigned on May 21, 1998 and the currency-board proposal was formally dead, the rupiah had already done most of its damage. The IMF and the new Habibie government converged on a free-float regime, and the rupiah eventually touched its weakest point around 16,000-17,000 per dollar before stabilizing into the 7,000-10,000 range over the second half of the year.
At the execution layer, the free float created a different problem: counterparty risk. Settlement on offshore rupiah trades became uncertain because domestic banking infrastructure was under strain. Banks that had been counterparties in January 1998 were not the same banks — legally or operationally — by August. Some had been recapitalized through the bank restructuring agency (IBRA), some had been merged, some had failed. Forward contracts written six months earlier became reconciliation puzzles. This is the part of the story that the macro narratives skip: a currency does not just trade in a chart, it settles into accounts at specific institutions, and when those institutions move under your trade the position you thought you had is not the position you actually have.
What Does the Episode Tell Us About Currency-Board Proposals Made Under Duress?
The lesson the orthodox literature draws is that currency boards are credible only when installed from a position of strength — when the reserves comfortably cover the base, when the political institutions can enforce discipline, when the corporate sector is not sitting on a wall of unhedged foreign liabilities. By that standard, February 1998 Indonesia was the wrong moment.
The lesson the heterodox literature draws is different. It is that any monetary regime imposed by a credit-providing institution on a borrowing country is, in the end, a political artifact. The IMF program was no less political than Hanke's board would have been; it was simply Washington's politics rather than Jakarta's. The rupiah's path through 1998 was shaped as much by the rejection of the board as by the absence of one.
The question that the archive leaves genuinely unanswered — and we mean unanswered, not rhetorical — is whether a credibly-backed board, installed in late 1997 before the panic phase, could have anchored the corporate sector's hedging behavior enough to prevent the cascade. If you have read the Bank Indonesia internal memoranda from that window, or if you traded the offshore rupiah through the period and remember what actually moved the bid, write.
FAQ
Why is the 1998 Indonesian currency-board proposal still studied today?
The episode sits at the intersection of two debates that remain live: whether hard pegs can stabilize a currency in a balance-of-payments panic, and whether IMF program design adequately distinguishes between confidence crises and balance-sheet crises. The proposal's rejection, combined with the rupiah's subsequent path to 16,000-plus, makes it a counterfactual that academics and policymakers continue to relitigate — particularly when newer crisis episodes (Argentina 2001, Turkey 2018, Sri Lanka 2022) raise similar questions about regime choice under duress.
How was Indonesia's situation different from Argentina's convertibility regime?
Argentina installed its currency board in April 1991 from relative calm, with a multi-year buildup of political consensus and a reserve position that credibly covered the narrow base at the announced rate of one peso per dollar. Indonesia would have been installing the same architecture during an active panic, with reserves that did not cover the base at any politically defensible rate, and without the institutional preconditions Argentina had spent two years assembling. Both regimes eventually collapsed, but the launch conditions were not comparable.
Did any operator infrastructure survive the rupiah crisis intact?
The institutional dealing desks at the bulge-bracket banks largely continued operating through the crisis, though counterparty lines to Indonesian banks were repeatedly redrawn. Refco was active in the futures and currency channels and continued through the period — its own collapse would come in 2005 for unrelated reconciliation reasons. The modern retail-FX brokerage layer (FXCM, the platforms that built around MT4) was not yet structurally present in 1998; most of those firms were founded later and entered Asia-Pacific markets in the 2000s.
Was Steve Hanke's involvement the reason the proposal failed?
No, though it was a contributing political friction. The proposal would have faced the same arithmetic challenge — insufficient reserves at the proposed rate — regardless of who designed it. Hanke's involvement made the rejection easier to frame in Washington, because it positioned the plan as an external academic intervention rather than an Indonesian government initiative. But the core IMF and Treasury objections were structural, not personality-driven.
What rate did the rupiah actually settle at after the free float?
The currency touched its weakest published levels around 16,000-17,000 per dollar in mid-1998 during the most acute political and banking-sector stress, then recovered into a 7,000-10,000 range as the Habibie government stabilized and the IMF program was reset. The rupiah did not return to pre-crisis levels at any point in the subsequent decade. The post-crisis trading range itself reflected a permanent repricing rather than a temporary overshoot.
Were there any winners on the institutional side during the rupiah's collapse?
The desks that had structurally short-rupiah positioning entering the second half of 1997 — primarily those expressing a view through forwards or NDF structures — did well. The corporate borrowers who had hedged their dollar liabilities (a minority) survived. The pattern, common across emerging-market crises, was that gains concentrated with offshore institutional participants who had the credit lines and settlement infrastructure to hold and unwind positions, while losses concentrated among domestic unhedged corporates and the depositors caught in the bank-closure phase.