Commentary · Financial System Leverage & Risk

A Dispatch From Jakarta, Twenty-Nine Years Late

Bank Indonesia is about to defend the rupiah with the same tool it reached for in 1997. The tool still works. This time the weather around it is worse.

BY MICHAEL WILSON · OCTOBER 20, 2026

Bank Indonesia will raise rates again. It has no real alternative. The rupiah is fixing at 18,050 to the dollar, within reach of June's record low of 18,190.¹˒² The policy rate already stands at 5.75% after 100 basis points of hikes since May, and the board meets tomorrow with forecasters split between a quarter-point and a half-point move.² A downgrade threat from MSCI has turned passive index money from a steady buyer of Indonesian assets into a scheduled seller. A central bank facing a scheduled seller has two choices: spend reserves or raise the price of leaving. It will do some of both. Defending the currency will cost domestic demand, and Indonesia will pay that cost.

Penblock covered the headline this morning. This column is about what sits behind it.

The cable I would have sent

John Paton Davies Jr. was a Foreign Service officer in wartime China. His dispatches from Chungking told Washington something it did not want to hear: the government it had chosen to back was hollow, its army was not fighting, and the people around it had stopped believing in it. He described the situation as it was, said where it was going, and put his name to it. He was right, and he was driven out of the State Department for it.

I keep his dispatches in mind because they model a discipline that financial commentary mostly lacks. Lead with the judgment. Separate what is happening from what officials say is happening. Say once what you think will occur, and why.

So here is the judgment. Indonesia in October 2026 is in better shape than Indonesia in 1997. The world around Indonesia is in worse shape than the world of 1997. The second fact matters more than the first.

The market already sees the parallel. On October 16 the Jakarta Composite fell about 11.6% in a single session, after a trading halt at the open, in what was reported as its worst day since January 1998.³˒⁴ It fell another 3.6% the following Monday.⁵

What 1997 actually was

The Asian Financial Crisis is remembered as a currency crisis. Underneath, it was a balance-sheet crisis that a currency crisis exposed.

Thailand floated the baht on July 2, 1997, after spending most of its reserves defending a peg. Investors then looked at Indonesia and saw the same structure: a currency held stable by policy, and a corporate sector that had borrowed heavily in dollars on the assumption that the stability would last. When the rupiah began to fall, every dollar debt grew in local terms. Companies scrambled to buy dollars to hedge, which pushed the rupiah down further. The rupiah went from roughly 2,400 to the dollar in mid-1997 to beyond 15,000 within a year.

The policy response made it worse. The IMF program that autumn demanded tight money and the abrupt closure of sixteen banks. Depositors concluded that no bank was safe and ran on all of them. Bank Indonesia's policy rate went to levels near 70 percent. The economy contracted by about 13 percent in 1998. In May of that year, after riots in Jakarta, Suharto resigned after three decades in power. The crisis began in Bangkok, but it was decided in Jakarta.

Three lessons from that sequence still apply:

Currency crises start in the balance sheets. The exchange rate is where the damage shows up. The cause is whoever owes money in a currency they do not earn.
Rate hikes defend a currency only when the economy can survive them. Past a certain point, a higher rate signals desperation and accelerates the exit it was meant to stop.
Markets punish opacity faster than they punish weakness. Indonesia's 1997 problem was partly economic and partly a belief that the true state of its banks and conglomerates was being hidden. Once that belief took hold, no official figure could restore confidence.

What is different now

An honest dispatch records the good news alongside the bad.

The rupiah floats. There is no peg to defend to the last dollar of reserves, so the adjustment happens through the price instead of through a sudden collapse. Bank Indonesia held $146.3 billion in reserves at the end of September, enough to cover 5.3 months of imports, well above the usual three-month adequacy benchmark.⁶ The banking system is better capitalized and better supervised. Inflation is 3.28%, inside the central bank's target band.⁷ Indonesia also has a legal deficit ceiling of 3 percent of GDP, written into law after 1998 as a direct answer to that crisis.

These are real defenses. They are the reason this is a stress event today and a solvency event only in the worst case.

What rhymes

The MSCI warning is the part of this story that should make anyone who lived through 1997 sit up.

In 1997 the hot money was bank loans and portfolio flows that could leave overnight. In 2026 a large share of it is passive: index funds and ETFs that own Indonesian stocks because an index provider says they should. That money does not do research and does not panic. It also does not stay. On October 16 MSCI proposed moving Indonesia out of its emerging-markets index to "standalone" status, and said it expects to proceed unless the country shows clear evidence of material governance improvements.⁸ When an index provider reclassifies, the selling happens on a known date at a known size, and every active investor in the world can see it coming. Front-running a forced seller is the easiest trade in finance. Foreign investors had already sold a net $4.88 billion of Indonesian stocks this year before the announcement landed.²

The reason for MSCI's warning should also sound familiar. Its concerns are opaque shareholding structures, suspected coordinated trading and weak enforcement, with free float at the center.⁸˒⁹ In plain terms: whether the shares that trade represent the real company, and whether outsiders can see who controls what. In 1997 the term was crony capitalism. Today the language is more polite. Both describe a market where outsiders worry they are not being told the whole story, and that worry sets the price.

There is a second rhyme in Jakarta. Over the past year investors have raised questions about the independence of Bank Indonesia and about the government's commitment to fiscal discipline, as large new state spending programs and a sovereign investment vehicle have taken shape. Each of these may be defensible on its own terms. Together, they mean the central bank is defending the currency while markets question whether it is free to do so. In 1997, credibility was the asset that ran out first. It is the asset to watch now.

Why the outside world makes this harder

In 1997 the crisis was in the periphery and the center was strong. The United States was in the middle of a long expansion with falling inflation and a balanced budget in sight. When the contagion reached Russia and then Long-Term Capital Management in 1998, the Federal Reserve cut rates three times and the system stabilized. The center had room to rescue the periphery.

The center does not have that room today. Three pressures are converging on it at once.

Oil. The war with Iran has delivered an oil shock. Brent is trading around $108 to $111 a barrel.¹⁰˒¹¹ Indonesia's own crude benchmark averaged $112.81 in September, up from $89.43 in August, against a 2026 budget built on $70 oil.⁷˒¹² Indonesia has been a net oil importer for two decades, and the government has promised not to raise subsidized fuel prices.⁷ The OECD estimates that holding the fuel price freeze at these oil prices adds about 0.6% of GDP to the deficit, pushing it to the 3% legal ceiling.¹³ A higher oil price widens the current account deficit, pulls dollars out of the country and strains the budget at the same moment Bank Indonesia is trying to hold the currency. In 1997 oil was cheap and falling. That was one of the few things that went right.

Treasuries. The U.S. 10-year yield is near 5.1%, after touching 5.31% on October 5. A year ago it was under 4%.¹⁴ Yields are rising because inflation has not been beaten and the federal debt keeps growing. Higher Treasury yields raise the return investors can earn without leaving home, which raises the price every emerging market must pay to hold its capital. It also means the Fed cannot respond to stress abroad the way it did in 1998. A central bank fighting inflation at home does not cut rates to calm Jakarta.

The AI trade. A large share of global equity gains in recent years has concentrated in a small group of U.S. companies tied to artificial intelligence. Valuations there assume spending and profits that have not yet arrived. When a concentrated trade like that unwinds, the losses force selling across everything investors own. Emerging-market assets are usually sold first, because they are the easiest to sell and the hardest to defend in an investment committee. Indonesia's stock market would absorb an AI correction it had no part in creating.

Each of these pressures is manageable alone. The danger is the combination: an emerging market under a scheduled outflow, an oil bill rising, a global risk-free rate climbing, and a world center that cannot cut rates and may soon be selling assets to cover its own losses.

What I am watching

Davies ended his reports with what Washington should look for next. Here is my list.

Tomorrow's decision and what comes with it. A measured hike paired with clear fiscal commitments buys time. An emergency hike with no fiscal message reads as panic.
Reserves spent per week. The rate decision is the announcement. The monthly reserve figure shows how hard the defense actually is. Watch whether the October number falls well below September's $146.3 billion.
The MSCI calendar. The consultation runs to April 30, 2027, with a final decision due by June 30.⁸ That gives Jakarta six months to show real governance changes, such as the planned doubling of the minimum free float to 15%.⁹ It also gives the market six months to front-run the outcome.
Dollar debt of Indonesian companies. This was the fuse in 1997. Watch for corporate issuers that cannot roll over foreign-currency bonds.
The neighbors. Contagion moves along trade and funding links. If the Philippine peso, Thai baht and Korean won begin to move with the rupiah, this is no longer an Indonesian story.
Oil and the 10-year Treasury together. Either one rising is a headwind. Both rising in the same week is the scenario this column is about.

The judgment, restated

Indonesia is not about to repeat 1998. Its defenses are real, and a floating currency with large reserves bends where the old peg broke. The threat is a slow squeeze: rates held high for longer than the economy wants, growth sacrificed to protect the currency, and a government tempted to blame foreign index providers instead of fixing the transparency problems that drew the warning.

The larger lesson of 1997 is about the gap between what a market is told and what it believes. When that gap opens, the official numbers stop mattering. Indonesia can close it by addressing MSCI's concerns directly and by leaving no doubt that its central bank is free to do its job. If it does, this will be remembered as a hard year. If it does not, the comparisons to 1997 will become more than a column.

Davies was punished for reporting accurately on a government that Washington had chosen to believe in. The warning did not cause the outcome. The outcome came because the warning was ignored.

This is author commentary and reflects the writer's personal views. It is not investment advice and does not recommend buying, selling or holding any security or currency. Market figures are as of October 19–20, 2026.

Sources

Relevant to:Financial Services & Banking

Signed commentary reflects the author's view, not Penblock's news reporting. It is not investment advice. Disclaimer