Hyperinflation is inflation that has become extremely high and keeps accelerating, not just one bad month of price rises. There is no single official threshold that every institution uses. Economist Phillip Cagan's classic 1956 study defines hyperinflation as beginning in the month the monthly rise in prices exceeds 50 percent, a cutoff his own paper calls purely arbitrary. The Federal Reserve Bank of Cleveland's explainer video on hyperinflation uses the same 50 percent a month line, noting that works out to more than 12,000 percent over a year. Accountants use a different test built for a different job: the international accounting standard IAS 29 says hyperinflation is indicated by factors such as cumulative inflation over three years of around 100 percent or more, among other signs, for deciding when a company must restate its financial statements. This page explains the common definitions, what causes hyperinflation, why it tends to feed on itself, how hyperinflations have ended in the past, named historical episodes exactly as sources state them, and a hypothetical worked example. It is a natural follow on to the velocity of money and fiscal dominance explainers on this site. It is educational only and is not investment, financial, tax or legal advice, and it makes no forecast about any current country's inflation.

Is there an official threshold for hyperinflation?

No single official number exists. Three reliable sources describe the idea three different ways, and the differences are not a mistake, they measure different things for different purposes.

  • Cagan's monthly rule (1956). In "The Monetary Dynamics of Hyperinflation", Phillip Cagan writes: "I shall define hyperinflations as beginning in the month the rise in prices exceeds 50 per cent and as ending in the month before the monthly rise in prices drops below that amount and stays below for at least a year." His own footnote adds that "the definition is purely arbitrary but serves the purposes of this study satisfactorily." The Federal Reserve Bank of Cleveland's video on hyperinflation uses the same rule of thumb, adding that 50 percent a month works out to more than 12,000 percent over a year, and notes that in the most extreme cases prices have doubled within days.
  • No official number, FRED Blog. The St. Louis Fed's FRED Blog post on hyperinflation puts it plainly: "There's no official threshold for 'very high'." The post simply points to clear cases, led by Zimbabwe, where FRED's own data show an inflation rate of 24,000 percent in a single year, meaning prices were multiplied by 241 times over.
  • An accounting trigger, IAS 29. The international accounting standard IAS 29, Financial Reporting in Hyperinflationary Economies, sets no single number either. It says hyperinflation is indicated by factors such as prices, interest and wages linked to a price index, and cumulative inflation over three years of around 100 percent or more. That test exists to tell a company when it must restate its financial statements in a current unit of currency, a different job from describing when economists reach for the word.

In short, Cagan's rule is the one most often quoted in economics classes and matches the Cleveland Fed's rule of thumb, while IAS 29's test is built for accountants, not students of monetary economics. All three sources agree hyperinflation means prices rising so fast that ordinary planning and contracts break down.

What causes hyperinflation?

Hyperinflations share one root cause rather than a sudden shock. Economist Michael K. Salemi, writing the hyperinflation entry for Econlib's Concise Encyclopedia of Economics, states that "hyperinflations are caused by extremely rapid growth in the supply of 'paper' money" and that "they occur when the monetary and fiscal authorities of a nation regularly issue large quantities of money to pay for a large stream of government expenditures." Put plainly, a government that cannot or will not raise enough through taxes or ordinary borrowing turns to the money printing press instead, which is the monetary policy side this site's fiscal dominance explainer covers. The FRED Blog's post on hyperinflation describes the same mechanism from the central bank's seat: hyperinflation generally happens "because the government takes over the central bank and finances its operations by printing money." Whichever money measure a reader has in mind, the M1 and M2 explainer on this site is about the much smaller, ordinary month to month changes in the money stock, a world away from the scale of money creation behind a hyperinflation.

The self reinforcing loop: why hyperinflation accelerates

A single burst of money creation does not by itself explain why prices can rise faster and faster. The missing piece is what people do with the money they hold. Salemi's Econlib article explains that once households expect prices to keep rising, they hold a lower real quantity of money, the ratio of money held to the price level, because holding cash becomes an implicit tax on its own: "they avoid the inflation tax by holding more of their wealth in the form of physical commodities. As they buy these commodities, prices rise higher and inflation increases." Cagan's 1956 paper, the first systematic study of this pattern, is built entirely around that same real cash balance ratio, tracking its sharp swings across seven historical hyperinflations in its data tables.

This is the same relationship covered on this site's velocity of money explainer, using the equation of exchange MV = PY, where M is the money stock, V is velocity, P is the price level and Y is real output. As expected inflation rises, people try to spend money faster rather than hold onto it, which raises velocity V. For a given stock of money M, a higher V pushes up PY, which in practice mostly shows up as a higher P once real output Y cannot keep pace, feeding inflation expectations further still. Salemi describes the result as a wasteful game people play trying to spend newly received money before it loses more value, and he calls hyperinflation self perpetuating: the government sees that its newly printed money buys less, so it prints still more, and the public tries to spend what it receives even faster than before. Try the velocity of money calculator to see how a change in any one of these variables moves the others in the identity.

How do hyperinflations end?

Econlib's Salemi describes the standard answer as a credible government commitment to stop the rapid money creation. He points to Germany's 1923 monetary reform as the case most often cited: the government introduced a new currency, the rentenmark, backed by a pledge that it could be converted on demand into a bond with a set value in gold, which Salemi reads as, in effect, a public promise to end the flood of new money. A second view he describes, associated with economist Thomas Sargent, holds that a credible stop usually needs two promises together, not one: an end to rapid money growth and a believable plan to bring the government's budget into balance, since an unbalanced budget is the reason the government turned to the printing press in the first place. Sargent judged the 1923 German reform successful partly because it also created an independent central bank able to refuse to finance the deficit, alongside higher taxes and lower government spending.

Other episodes have ended differently. The FRED Blog notes that Zimbabwe's hyperinflation was broken only by "a radical change of practice": abandoning the use of its own local currency altogether, a step economists call dollarization when a country turns to a foreign currency such as the US dollar instead.

Named historical examples

The Federal Reserve Bank of Cleveland's video on hyperinflation names four episodes without giving country by country figures: "a famous example of hyperinflation occurred in Germany between World War I and World War II. More recent examples include Venezuela starting in 2017, Zimbabwe in the 2000's and Yugoslavia in the 1990's." The FRED Blog adds one dated figure for part of Zimbabwe's episode: an inflation rate of 24,000 percent in a single year, meaning prices were multiplied by 241 times over, noting that Zimbabwe stopped reporting an inflation rate altogether in 2009 because the true figure likely ran even higher by then. This page does not add numbers for any country beyond what these two sources state, and it makes no claim about any country's inflation rate today.

Hyperinflation vs ordinary inflation, high inflation and stagflation

Hyperinflation sits at the extreme end of a scale that most of macroeconomics is not about. Ordinary inflation, the kind targeted by the Federal Reserve and the Bank of England, is covered in the inflation targeting and the 2 percent goal explainer, built around a target near 2 percent a year, nowhere close to 50 percent a month. The opposite end of the price scale, falling prices and a slowdown in the inflation rate, is covered in deflation vs disinflation. Stagflation is a different problem again, weak growth alongside unusually high inflation, which is awkward for policy, but it is not the monetary breakdown behind hyperinflation. Whichever of these a reader is studying, the CPI vs PCE explainer covers how the price indexes behind an ordinary inflation rate are actually built, and the real interest rates explainer covers why very high or unstable inflation makes a nominal interest rate hard to interpret on its own.

A hypothetical worked example

Example (illustrative)

These numbers are invented to show the arithmetic at exactly Cagan's 50 percent a month boundary. They are not a forecast or data about any real currency or country. If prices rise 50 percent every month, a $1.00 item costs $1.50 after one month and 1.5 squared, or $2.25, after two months. After 12 months the item costs 1.5 to the power 12, which works out to $129.75 (Econlib's own worked example, at the same 50 percent a month rate, rounds this to $130). That is an increase of about 12,875 percent over the year, even though the month to month rate never rose above 50 percent. Written as a constant daily rate, 50 percent a month compounds out to 1.5 to the power of 1 divided by 30, minus 1, assuming a 30 day month, which is about 1.36 percent a day. At that pace, prices double roughly every 1.71 months, or about 51 days, found from the natural log of 2 divided by the natural log of 1.5. Real hyperinflations are rarely this tidy, since the monthly rate changes throughout the episode, and the Cleveland Fed notes that in the worst stretches prices have doubled within days, far faster than this boundary case.

Common exam mistakes

  • Confusing a high price level with a high inflation rate. A country can have had very high prices for years without current hyperinflation; what matters for the definitions on this page is the rate of change, not the level.
  • Thinking a single supply shock can cause hyperinflation on its own. Econlib's Salemi is explicit that no single shock, however severe, explains sustained and continuously rapid price growth; hyperinflations need ongoing, rapid money creation to finance government spending.
  • Treating 50 percent a month as an official, universal legal threshold. It is Cagan's own research choice, which his footnote calls purely arbitrary, and the IAS 29 accounting test is a separate, unrelated indicator built for financial reporting.
  • Assuming hyperinflation ends the moment money printing stops. Sargent's view, described in the Econlib source above, is that some episodes also needed a credible fiscal reform, and Zimbabwe's episode ended only once the country dropped its own currency.

Related reading

Sources

Frequently asked questions

What is hyperinflation in plain terms?

Hyperinflation is inflation that has become extremely high and keeps accelerating, rather than a single bad month of price rises. The Federal Reserve Bank of Cleveland describes it as prices rising rapidly and generally at an increasing pace, usually associated with social upheaval and civil unrest.

Is there one official threshold for hyperinflation?

No single official number exists. Economist Phillip Cagan's 1956 study defines it as beginning the month prices rise more than 50 percent, a line his own footnote calls purely arbitrary. The Federal Reserve Bank of Cleveland uses the same 50 percent a month rule of thumb. The FRED Blog says plainly there is no official threshold for very high inflation. The accounting standard IAS 29 uses a different, unrelated indicator, cumulative inflation over three years of around 100 percent or more, built for financial reporting rather than for defining the economic term.

What actually causes hyperinflation?

Sustained, very rapid growth in the money supply used to finance large government spending. Econlib's entry on hyperinflation states that hyperinflations are caused by extremely rapid growth in the supply of paper money, occurring when monetary and fiscal authorities regularly issue large quantities of money to pay for a large stream of government expenditures.

How have hyperinflations ended in the past?

Econlib describes the standard answer as a credible government commitment to stop rapid money creation, citing Germany's 1923 currency reform. Economist Thomas Sargent's view, also described there, is that a credible stop usually needs fiscal reform too, a believable plan to balance the budget, alongside the monetary pledge. The FRED Blog notes that Zimbabwe's episode ended only once the country abandoned the use of its own local currency.

Is hyperinflation just very high ordinary inflation?

Not quite. Ordinary inflation, including the 2 percent targets most central banks aim for, is a gradual process managed with interest rates. Hyperinflation is a monetary breakdown, usually driven by a government financing its spending through money creation, and it feeds on itself once people expect prices to keep rising.