Brazil’s Hyperinflation and the Real Plan: What Six Currencies in Eight Years Actually Fixed
For four months in 1994, a can of soda on a São Paulo shelf carried two prices on the same tag: one in cruzeiros reais, the currency actually changing hands, and one in URV — a unit that existed only on paper, recalculated daily, that nobody could hold in a wallet. That dual pricing is how Brazil ended a hyperinflation that had already outlasted five previous currencies.
Our own money-supply series confirms what the URV period was fighting. Brazil’s money supply grew 3,280.7% in 1993, the peak year of the run-up, by our own calculation from the raw IMF series. Brazil’s series runs continuously from 1960 through May 2026 — the whole Brazil hyperinflation episode and the recovery after it sit on one line, with one change of source, in December 2001. We mark that seam wherever it touches a number.
Specs — Brazil broad money
| Indicator | Broad money (M2/M3 equivalent) |
| Latest value | R$15.63 trillion ($2.80T USD) |
| Observation date | 2026-05 |
| Year-on-year growth | 11.0% (12 months to 2026-05) |
| 10-year CAGR | 11.1% (through 2026-05) |
| Doubling time at that rate | 6.6 years |
| 1993 peak run-up year | +3,280.7% money-supply growth (our calc) |
| Post-Plan low, 1998 | +12.0% — before the 1999-2000 reversal below |
| Our coverage | 1960 → 2026-05 · 335 observations (World Bank annual 1960-2000, then IMF monthly from 2001-12) |
| Source | IMF MFS_MA/BM_MAI · USD via World Bank PA.NUS.FCRF (period average) · source seam 2001-12 |
Six currencies, eight years, and the zeros each one cut
| Currency | Introduced | Conversion | Zeros cut |
|---|---|---|---|
| Cruzeiro | in use since 1970 | — | — |
| Cruzado | Feb 1986 | 1 Cruzado = 1,000 Cruzeiros | 3 |
| Cruzado Novo | Jan 1989 | 1 Cruzado Novo = 1,000 Cruzados | 3 |
| Cruzeiro | Mar 1990 | 1 Cruzeiro = 1 Cruzado Novo | 0 (renamed only) |
| Cruzeiro Real | Aug 1993 | 1 Cruzeiro Real = 1,000 Cruzeiros | 3 |
| Real | Jul 1994 | 1 Real = 2,750 Cruzeiros Reais | not a round cut — see the URV below |
Six currency names circulated in those eight years, and only the sixth held. Money-supply growth ran high triple and quadruple digits nearly every year of the stretch, by our own calculation from the raw series: 289.2% (1986), 213.7% (1987), 2,733.0% (1988), 1,904.9% (1989), 640.9% (1990), 501.6% (1991), 1,749.4% (1992), 3,280.7% (1993). None of the first five redenominations touched that trajectory — chopping zeros off a currency’s face value doesn’t change what’s driving the number underneath it.
None of those cuts shows up in our series as a step down either. The World Bank backfill is stated in constant modern reais across the whole 1960-2000 stretch. So 3,280.7% is a growth rate, not the 2,750-to-1 conversion in disguise.
We hold Brazil’s money supply, not its consumer prices. For the price side we use somebody else’s figures and name them. Hanke and Krus, in World Hyperinflations (Cato Institute working paper, 2012; reprinted in the Routledge Handbook of Major Events in Economic History, 2013), date Brazil’s hyperinflation to four months: December 1989 through March 1990. They put the worst month at March 1990, when prices rose 82.4%. That clears Cagan’s 50%-a-month threshold, but only inside that short window. Across most of the eight years above, Brazilian inflation ran extraordinarily high and stayed under the bar. The money-supply figures in this section are ours. The price figures are theirs, and the two are not interchangeable.
The URV: the genuinely clever part nobody explains simply
Brazil’s fix didn’t start with a new currency. It started with a currency that wasn’t real. On 1 March 1994, the government introduced the Unidade Real de Valor (URV) — a unit of account, not something you could hold, pegged close to the US dollar and recalculated daily. Every price, wage and contract in the country was quoted in both cruzeiros reais and URV side by side for four months. People learned to think in a stable unit before that unit ever became actual money.
On 1 July 1994, the government retired the cruzeiro real and the URV together and issued the real, fixed at 2,750 cruzeiros reais to 1 real. Everyone had already been pricing in URV for months. The changeover didn’t reset expectations the way the five earlier redenominations had. The trick wasn’t the exchange rate. It was breaking the automatic re-indexing habit. Prices and wages had been re-adjusting to yesterday’s inflation the moment a new plan tried to freeze them. That habit had defeated every previous plan outright.
The Brazil hyperinflation aftermath: what our own series shows after 1994
Growth fell every single year for four years after the real launched: 44.3% in 1995, 31.0% in 1996, 17.2% in 1997, 12.0% in 1998 — the smoothest deceleration among the endings in our dataset over that stretch.
It did not hold in a straight line. Brazil floated the real in January 1999, under pressure from the Asian and Russian financial crises already hitting currencies worldwide. Money-supply growth reversed: 18.1% in 1999, then 19.7% in 2000. Both are our own calculation from clean World Bank backfill points, and both are a real reversal rather than an artifact of a currency swap.
We stop the sequence there, and the reason is worth publishing. Brazil’s series changes source at December 2001 — World Bank annual data through 2000, IMF monthly from 2001-12 — and a year-on-year rate computed across that seam measures the gap between two definitions as much as a year of growth. Our arithmetic across it returns 38.4%. We measured Brazil’s cross-source deviation at 20.9 percentage points, over half that figure, so it does not ship as a growth rate. Four years of clean deceleration is still not the same as a permanently solved problem, and the 1999-2000 pair says so on its own.
By 2026-05, the same series puts Brazil’s broad money at $2.80 trillion, growing 11.0% year-on-year and 11.1% on a 10-year average — a 6.6-year doubling time. That’s ordinary large-economy monetary growth today, not a hyperinflation-aftermath number. It just didn’t arrive by the straight line the first four post-Plan years promised.
The Brazil hyperinflation ending is one of four charted end-to-end in how hyperinflation actually ends, alongside Argentina’s currency board, Peru’s fiscal shock and Zimbabwe’s dollarization. See Brazil measured directly against Argentina’s relapse on our Brazil vs. Argentina comparison, and Brazil’s full monthly history on our country page.
FAQ
What ended the Brazil hyperinflation of the 1980s and early ’90s?
The Real Plan of 1994. It introduced a transitional unit of account, the URV, that re-indexed prices and wages for four months before the real itself launched on 1 July 1994 at 2,750 cruzeiros reais to 1 real — breaking the automatic re-indexing habit that had defeated five earlier currency changes.
Did Brazil’s Real Plan solve the problem permanently?
Not permanently, though it worked cleanly at first. Money-supply growth fell every year from 1995 (44.3%) to 1998 (12.0%), then reversed to 18.1% (1999) and 19.7% (2000) after Brazil floated the real under pressure from the Asian and Russian financial crises. We stop the sequence at 2000: our 2001 reading straddles a change of source and is not a usable growth rate.
How many times did Brazil change its currency?
Six currency names circulated between 1986 and 1994 alone — cruzeiro, cruzado, cruzado novo, cruzeiro (again), cruzeiro real, and finally the real. Only the sixth one held.
Is Brazil’s money supply still growing fast today?
By ordinary standards, yes; by hyperinflation standards, no. As of 2026-05, Brazil’s broad money is growing 11.0% year-on-year and 11.1% on a 10-year average — a 6.6-year doubling time, in line with a normal large economy rather than a hyperinflation aftermath.