Hyperinflation vs Inflation: What Is the Actual Difference?
Hyperinflation begins the month a country's price level rises by at least 50 percent, the threshold economist Philip Cagan set in his 1956 study The Monetary Dynamics of Hyperinflation. Everything below that line, including annual inflation of 20, 60, or even 100 percent, is inflation. It may be painful inflation, but it is a different economic regime with different mathematics and different banknotes.
Last updated: July 2026
Inflation and hyperinflation are separated by one specific number: 50 percent price growth in a single month, the threshold Philip Cagan established in 1956 and the rule the Hanke-Krus World Hyperinflation Table (Cato Institute) applies to every episode it documents. Under Cagan's rule an episode starts in the month the price level climbs at least 50 percent, and it ends when monthly inflation falls back below 50 percent and stays there for a full year (Hanke and Krus, "World Hyperinflations," Cato Working Paper, August 15, 2012). Sustained for a year, 50 percent monthly compounds to roughly 12,875 percent, about 1.4 percent every single day, with prices doubling roughly every 52 days. Hanke and Krus used that rule to build the most complete systematic catalog of the phenomenon, and it contains exactly 56 episodes in all of recorded history. That scarcity is why hyperinflation banknotes are a distinct collecting category rather than a footnote to world paper money.
Where exactly is the line between inflation and hyperinflation?
The line sits at 50 percent price growth in one month. Philip Cagan set it in 1956, and Steve Hanke and Nicholas Krus adopted it verbatim when they compiled the Hanke-Krus World Hyperinflation Table, describing it as "a price-level increase of at least 50% per month" (Hanke and Krus, Cato Working Paper, 2012).
What makes Cagan's rule useful is that it is a rule about months, not years. Almost every inflation figure the public sees is annual. Cagan's threshold ignores that framing and asks a sharper question: did prices rise by half in thirty days? That single reframing is why two countries with similar-sounding annual numbers can sit on opposite sides of the definition.
The rule has a second half that is often forgotten. Hanke and Krus write that under Cagan's definition, "when the monthly inflation rate drops below 50% and stays there for at least one year, the episode is said to end." A country that crosses the line in March and falls back in April is still in a hyperinflation episode until it has strung together twelve consecutive months underneath the threshold. That clause is what turns a chaotic run of monthly data into a countable episode with a start date and an end date.
Apply the compounding and the threshold stops being abstract. At exactly 50 percent per month, prices multiply by 1.5 twelve times in a year, which is a factor of about 130. Expressed as an annual rate that is roughly 12,875 percent. Expressed as a daily rate it is about 1.4 percent, every day, including weekends. Those figures are simple arithmetic on the threshold itself rather than a measurement of any particular country, but they explain why the number was chosen: 50 percent a month is the point at which the calendar a household uses to think about money stops being the year and starts being the week.
Ordinary inflation, high inflation, and hyperinflation are three different regimes
These are not three points on one smooth scale. They are three regimes with different anchors, different institutional responses, and different consequences for physical currency. Ordinary inflation is a policy target, high inflation is an accounting problem, and hyperinflation is a currency's terminal phase.
At the ordinary end, inflation is deliberately produced and deliberately capped. The Federal Open Market Committee states that "inflation of 2 percent over the longer run, as measured by the annual change in the price index for personal consumption expenditures, is most consistent with the Federal Reserve's mandate for maximum employment and price stability" (FOMC Statement on Longer-Run Goals and Monetary Policy Strategy, Federal Reserve). A 2 percent annual target is roughly 0.17 percent a month. Cagan's threshold is roughly three hundred times that pace.
The middle regime has its own formal marker, and it comes from accounting rather than monetary economics. IAS 29, the international standard titled Financial Reporting in Hyperinflationary Economies, tells companies to restate their financial statements when an economy shows warning signs including "cumulative inflation over three years of around 100 per cent or more" and prices, interest, and wages becoming "linked to a price index" (IFRS Foundation, IAS 29). That is a far lower bar than Cagan's. An economy can trip the accounting standard while never coming close to 50 percent in any single month, which is precisely what makes the two definitions complementary rather than contradictory.
| Regime | Defining benchmark | Roughly, per month | Who defines it | What happens to banknotes |
|---|---|---|---|---|
| Ordinary inflation | 2 percent a year (US longer-run objective) | About 0.17 percent | Federal Reserve, FOMC Statement on Longer-Run Goals | Denomination structure unchanged for decades |
| High inflation | Cumulative inflation near or above 100 percent over three years | About 1.9 percent | IFRS Foundation, IAS 29 | New high denominations added, old low ones quietly retired |
| Hyperinflation | 50 percent or more in a single month | 50 percent or more | Philip Cagan (1956); Hanke-Krus World Hyperinflation Table, Cato Institute | Zeros added faster than presses can run, then full redenomination |
Monthly equivalents for the first two rows are compounded from the stated annual and three-year benchmarks. They are arithmetic on the published thresholds, not measured rates for any specific country.
Why did Cagan choose 50 percent a month rather than some other number?
Because 50 percent a month is roughly where money stops working as a store of value and becomes a hot potato. At that pace prices double in about 52 days, and holding cash for a month costs a third of its purchasing power, so people begin spending on receipt rather than saving.
Hanke and Krus are candid that the literature before them was "riddled with a variety of definitions, and more often than not, they are vague and ill-defined." They chose Cagan's threshold because it is precise, replicable, and already widely accepted. A rule that applies identically to Weimar Germany in 1923 and to Zimbabwe in 2008 beats a more intuitive rule that different researchers apply differently.
They also note that Cagan did not always follow his own rule. Hanke and Krus point out that Cagan "selectively excludes Germany's 1920 case of hyperinflation, presumably because of its short duration (one month)." Applied strictly, that single month of 56.9 percent inflation in January 1920 qualifies, which is why Germany appears twice in the Hanke-Krus table: once for the brief 1920 episode and once for the famous August 1922 to December 1923 collapse that produced the notes in our Germany hyperinflation guide. Hungary appears twice as well, for the 1923 to 1924 crown episode and the far more extreme 1945 to 1946 pengő collapse.
The behavioral break is the real justification. Above the threshold, contracts shorten, wages get paid daily, and shopkeepers reprice mid-morning. Below it, an annual salary negotiation still means something. Our banknote glossary covers the vocabulary that appears on notes issued during that break, from overprints to emergency scrip.
How does the Hanke-Krus table decide what counts?
The Hanke-Krus World Hyperinflation Table documents 56 episodes, applies Cagan's 50 percent monthly rule strictly, and records the start date, end date, peak month, peak monthly rate, equivalent daily rate, price-doubling time, circulating currency, and price index used for every one of them.
Hanke and Krus preferred consumer price indices because they "best reflect price changes experienced by the final consumer," falling back to wholesale indices when consumer data did not exist. In the most extreme cases even that was impossible, so they used proxies. Greece's rate was reconstructed from the drachma price of the gold sovereign. Zimbabwe's was derived from an implied exchange rate calculated from a stock traded simultaneously in Harare and London. That last choice matters: Zimbabwe's official consumer price index stopped in July 2008, and the peak actually arrived roughly three and a half months later, in mid-November 2008.
The table is also disciplined about what it leaves out. The authors identified a probable North Korean hyperinflation around 2009 to 2011 but excluded it because "the available North Korean data did not meet our minimum quality standards," noting that including it would have brought the total to 57. That willingness to omit a case rather than publish a shaky number is the reason the table is treated as the reference standard.
Two corrections in the table are worth knowing if you collect these notes. Hanke and Krus found that the world's first hyperinflation, in revolutionary France, peaked with the mandat rather than the assignat, "generating a monthly inflation rate of 304% in mid-August 1796." And they documented a largely unreported Philippine episode under Japanese occupation, peaking at 60.0 percent in January 1944 with Japanese war notes in circulation. Both are collectible today: see our guides to French assignats and Japanese invasion money.
What do the boundary cases look like?
The mildest episode in the entire Hanke-Krus table is Taiwan in February 1947, at 50.8 percent for the month, barely over Cagan's line. The most severe is Hungary in July 1946, at 4.19 × 1016 percent. Both are hyperinflations by the same rule, which shows how much range the single definition covers.
| Location | Peak month | Highest monthly inflation | Equivalent daily rate | Time for prices to double | Currency |
|---|---|---|---|---|---|
| Hungary | July 1946 | 4.19 × 1016 percent | 207 percent | 15.0 hours | Pengő |
| Zimbabwe | Mid-November 2008 | 7.96 × 1010 percent | 98.0 percent | 24.7 hours | Dollar |
| Yugoslavia | January 1994 | 313,000,000 percent | 64.6 percent | 1.41 days | Dinar |
| Germany | October 1923 | 29,500 percent | 20.9 percent | 3.70 days | Papiermark |
| Greece | October 1944 | 13,800 percent | 17.9 percent | 4.27 days | Drachma |
| France | Mid-August 1796 | 304 percent | 4.77 percent | 15.1 days | Mandat |
| Argentina | July 1989 | 197 percent | 3.69 percent | 19.4 days | Austral |
| Brazil | March 1990 | 82.4 percent | 2.02 percent | 35.1 days | Cruzado & Cruzeiro |
| Philippines | January 1944 | 60.0 percent | 1.58 percent | 44.9 days | Japanese war notes |
| Germany | January 1920 | 56.9 percent | 1.51 percent | 46.8 days | Papiermark |
| Taiwan | February 1947 | 50.8 percent | 1.38 percent | 51.4 days | Taipi |
All figures from the Hanke-Krus World Hyperinflation Table in Hanke, Steve H., and Nicholas Krus, "World Hyperinflations," Cato Working Paper, August 15, 2012. Taiwan February 1947 is the least severe of the 56 documented episodes; Hungary July 1946 is the most severe.
Notice how quickly the ladder steepens. Taiwan's 50.8 percent month took nearly two months to double prices. Germany's October 1923 doubled them every 3.7 days, Zimbabwe's peak in under 25 hours, Hungary's in 15. There is more distance between the top and bottom of this table than between its bottom row and a quiet 2 percent economy. Our full severity ranking pairs the same comparison with the notes each crisis produced.
Modern high-inflation economies illustrate the other side of the line. Argentina posted triple-digit annual inflation in 2023 (INDEC), a genuinely severe crisis, without any single month reaching Cagan's threshold. Venezuela did cross it: Steve Hanke dated the onset to late 2016, when monthly inflation exceeded 50 percent for thirty consecutive days, and the IMF estimated 2018 inflation at 929,790 percent (IMF). The episode ended on the twelve-month rule, with the Central Bank of Venezuela and Venezuelan economists placing the exit in early 2022. Our Venezuela hyperinflation guide traces the three currencies that crisis passed through.
Why only hyperinflation prints the giant denominations collectors want
High inflation lets a treasury add a zero every few years. Hyperinflation forces it to add zeros faster than the presses can physically deliver them, which is the mechanism that produces trillion-denomination notes. Nearly every famous giant-denomination banknote in collecting comes from one of the 56 documented episodes.
Under ordinary or even high inflation, a central bank manages denominations gradually. It retires the smallest coin, introduces a larger note once a decade, and eventually runs a tidy redenomination that lops off three zeros. Under hyperinflation the arithmetic overwhelms that process. When prices double every four days, a note printed on Monday buys half as much by Friday, so the treasury skips past the next denomination and prints the one after it.
| Crisis | Peak month | Highest-denomination note issued |
|---|---|---|
| Hungary, pengő | July 1946 | 100 quintillion pengő |
| Zimbabwe, dollar | Mid-November 2008 | 100 trillion dollars P-91 |
| Germany, Papiermark | October 1923 | 100 trillion marks |
| Yugoslavia, dinar | January 1994 | 500 billion dinara |
| Greece, drachma | October 1944 | 100 billion drachmai |
Peak months from the Hanke-Krus World Hyperinflation Table, Cato Institute. Denominations follow the Standard Catalog of World Paper Money (Pick) listings for each currency.
This is why the definitional question matters to collectors and not just to economists. The 50 percent monthly threshold is, in practice, the threshold at which a country starts producing the paper that ends up in albums. Hungary's 100 quintillion pengő is the highest denomination ever put into circulation (Guinness World Records). Zimbabwe's 100 trillion dollar note is the modern icon, and the German 100 trillion mark is its Weimar ancestor. Our guide to the highest denomination banknote ever issued settles the ranking among them.
Because there have only ever been 56 episodes, and because most were resolved with a redenomination that withdrew and demonetized the old notes, the surviving supply is fixed and finite. That is the collecting case in one sentence. If you are starting out, our advice on how to collect world banknotes and the best hyperinflation banknotes to buy points you at the episodes with the best mix of story, availability, and eye appeal, and Pick catalog numbers explain how each issue is identified.
Frequently asked questions
What is the difference between inflation and hyperinflation?
Inflation is any sustained rise in the general price level, usually reported as an annual rate. Hyperinflation is a specific threshold within that: a price-level increase of at least 50 percent in a single month, per Philip Cagan's 1956 definition, which the Hanke-Krus World Hyperinflation Table (Cato Institute) applies to all 56 documented episodes. The distinction is about monthly speed, not annual size.
Who defined hyperinflation as 50 percent per month?
Economist Philip Cagan, in his 1956 study The Monetary Dynamics of Hyperinflation. Steve Hanke and Nicholas Krus adopted his rule without modification for the Hanke-Krus World Hyperinflation Table, describing it as "a price-level increase of at least 50% per month" and calling it the widely accepted standard (Hanke and Krus, Cato Working Paper, August 15, 2012). It remains the benchmark economists use today.
Is 100 percent annual inflation hyperinflation?
No. One hundred percent a year averages under 6 percent a month, far below Cagan's 50 percent monthly threshold. It is severe high inflation, and it does trigger the accounting standard IAS 29, which flags economies with "cumulative inflation over three years of around 100 per cent or more" (IFRS Foundation). But it is not hyperinflation, and it does not typically produce record denominations.
How does a hyperinflation officially end?
Under Cagan's rule, an episode ends only after monthly inflation drops below 50 percent and stays below it for at least one year (Hanke and Krus, Cato Working Paper, 2012). A single calm month is not enough. This twelve-month clause is why Germany and Hungary each appear twice in the Hanke-Krus table, as separate episodes rather than one long crisis.
Why do only hyperinflations produce trillion-denomination banknotes?
Because zeros get added faster than presses can keep up. At Germany's October 1923 peak, prices doubled every 3.70 days, and at Zimbabwe's November 2008 peak, every 24.7 hours (Hanke-Krus, Cato Institute). No treasury can manage denominations gradually at that pace, so it skips ahead by orders of magnitude. That is how the 100 trillion mark and the 100 trillion dollar notes came to exist.
Planet Banknote is a family-owned dealership in Sarasota, Florida, founded in 2021. Every note is sourced direct from mints, central banks, and authorized distributors, inspected through our Planet Banknote Verified process, and ships with a free Certificate of Authenticity. US orders ship free via USPS Priority, and every order includes a free bonus gift.
Now that the definition is settled, put it to work. Compare all 56 documented episodes side by side in Every Hyperinflation in History, Ranked, then read the crisis narratives for Zimbabwe, Yugoslavia, and Greece. When you are ready to hold one, browse our curated hyperinflation sets, the Zimbabwe banknotes category, or the Venezuela collection, and prices vary with inventory.