How far did the Singapore dollar fall in each crisis, and how fast did it come back?
The Singapore dollar fell 23.1% in the 1997 episode and between 5.0% and 11.8% in the four since. What separates it from the rest of the region is the return: four of the five episodes were recovered within seven months.
How far did the Singapore dollar fall in each crisis?
By 23.1% in the 1997 episode and by between 5.0% and 11.8% in the four since. The table below uses the monthly average of the Singapore dollar against the US dollar from a single series, measured from the base month to the highest monthly average in that window (BIS bilateral rates).
How far did the Singapore dollar fall in each crisis?
Depreciation
Shared linear scale from zero. Selected entries from the table below. Its dates, units, definitions and limitations still apply.
| Episode | Base month | Peak (monthly average) | Depreciation | Time taken |
|---|---|---|---|---|
| 1997 Asian financial crisis | 1.4271 (Jun 1997) | 1.7571 (Aug 1998) | +23.1% | 14 months |
| 2008 global financial crisis | 1.3681 (Jun 2008) | 1.5292 (Mar 2009) | +11.8% | 9 months |
| 2020 Covid | 1.3516 (Jan 2020) | 1.4246 (Apr 2020) | +5.4% | 3 months |
| 2022 tightening | 1.3510 (Jan 2022) | 1.4252 (Oct 2022) | +5.5% | 9 months |
| 2024–2026 | 1.2963 (Sep 2024) | 1.3611 (Jan 2025) | +5.0% | 4 months |
In the same 1997 episode the Indonesian rupiah fell 467.4% and the Thai baht 116.0%. By size of move, Singapore was among the least affected markets in the region.
All figures are monthly averages, not daily closes or intraday extremes, which are always further out during volatile periods. The base month is fixed at June 1997 for every market so the episodes can be compared, which means it does not line up with the dates on which individual countries changed policy. The site's calculators use a different exchange-rate series from this table, so decimals may differ slightly; do not mix the two series in one table.
The recovery is the part worth reading
Depreciation tells you how far the rate went. Recovery tells you how long it stayed there, and for this market the second number is the more distinctive one.
| Episode | Depreciation | Time to recover |
|---|---|---|
| 1997 Asian financial crisis | +23.1% | 114 months |
| 2008 global financial crisis | +11.8% | 17 months |
| 2020 Covid | +5.4% | 7 months |
| 2022 tightening | +5.5% | 3 months |
| 2024–2026 | +5.0% | 4 months |
Recovery is counted from the peak month to the first month in which the monthly average returned to the base-month level or below. Three of the last four episodes closed within seven months. That is the fastest set of returns among the nine markets compared in this series.
The exception is 1997. Recovery to the June 1997 level of 1.4271 took 114 months and arrived only in February 2008 — nine and a half years. A 23.1% move and a nine-and-a-half-year recovery belong to the same episode, which is the clearest reminder that size and duration answer different questions.
As of August 2026 the monthly average is 1.2763, stronger than the base month of all five episodes.
Was Singapore unusual in 1997?
Yes, in size. The table below measures from June 1997 to each currency’s own peak using the same method.
| Currency | Depreciation from Jun 1997 | Time taken |
|---|---|---|
| Indonesian rupiah | +467.4% | 13 months |
| Thai baht | +116.0% | 7 months |
| Malaysian ringgit | +75.2% | 7 months |
| Philippine peso | +66.0% | 15 months |
| Taiwan dollar | +24.5% | 14 months |
| Singapore dollar | +23.1% | 14 months |
| Hong Kong dollar | +0.4% | 30 months |
For a reference outside the region, the Korean won fell 90.4% in seven months. Singapore and Taiwan recorded almost the same move over the same fourteen months, but Taiwan then needed 274 months to return to its June 1997 level against Singapore’s 114.
Singapore manages the exchange rate rather than an interest rate, guiding the trade-weighted Singapore dollar within a policy band ([MAS](https://www.mas.gov.sg/monetary-policy)). Because the managed variable is a basket rather than the US dollar alone, the bilateral rate used in these tables is one component of that framework and not the policy target itself. This paragraph interprets the arrangement; the figures in the tables are observations from the monthly exchange-rate series.
What to check when reading these numbers
Start with the basis of measurement. Monthly averages, daily closes and intraday extremes give three different answers for the same event. This page uses monthly averages throughout so that five episodes stay comparable.
Then check the base month. Change the base month and the percentage changes immediately. This table uses one set of base months for every market, which makes cross-market comparison possible but does not match each country’s own policy dates.
Finally, separate the exchange rate from purchasing power. A 23.1% depreciation does not mean domestic prices rose 23.1%. The exchange rate changes import prices first, and the pass-through into the consumer price index is partial and takes time.
Questions & answers
How much did the Singapore dollar fall in the 1997 Asian financial crisis?
On a monthly-average basis, 23.1%. The rate went from 1.4271 in June 1997 to 1.7571 in August 1998. Over the same period the Thai baht rose 116.0% and the Indonesian rupiah 467.4%, so Singapore was among the least affected markets in the region by size of move.
How long does the Singapore dollar usually take to recover?
In the four most recent episodes, between three and seventeen months. The 2022 episode recovered in three months, the 2024 episode in four, the 2020 episode in seven and the 2008 episode in seventeen. The 1997 episode is the exception, taking 114 months.
Is the Singapore dollar weaker now than before these episodes?
No. The August 2026 monthly average of 1.2763 is stronger than the base month of every one of the five episodes, the most recent of which was 1.2963 in September 2024. Its strongest monthly average since 1990 was 1.2090 in August 2011.
Why do the numbers here differ from the rates quoted in the news?
This page uses a single monthly-average series across the whole table so that five episodes can be compared on one basis. News reports usually quote daily closes or intraday extremes, which give more extreme figures during volatile periods.