The argument this series makes, in 28 steps. It is the same spine every module is built on, and no module states more of it than its own share.
When money stops holding its value people stop saving it: a wage is spent within hours of being paid, and at Zimbabwe's peak in 2008 prices roughly doubled every day.
The state runs out of digits before it runs out of will: notes are reissued with more zeros, the zeros are struck off in a redenomination, and within months they are back.
A hyperinflation is a bounded, dated episode rather than a degree of bad inflation, and it is short: Germany from August 1922 to December 1923, Zimbabwe from March 2007 to mid-November 2008, Venezuela from November 2017 to January 2022.
It has happened about fifty times, which makes it a mechanism rather than one country's disgrace.
There is no separate act called printing: a central bank creates money by buying something, so the question is never why it printed but what it was buying.
What it was buying, in all three cases, was its own government's deficit: money the state was spending that it had not raised.
Because the new money is created as the government's spending, it arrives as demand immediately rather than sitting anywhere.
Germany's inflation began in 1914, not 1923: the war was financed by borrowing and note issue rather than by taxation, and the mark had already lost most of its pre-war value before the hyperinflation proper began.
Reparations then took nearly half the German budget, and when France and Belgium occupied the Ruhr in January 1923 Germany paid its workers to stop working and created the money to pay them.
Zimbabwe's central bank was itself the spender, funding government programmes directly, so there was no gap between the deficit and its financing to be crossed.
Farm output roughly halved after land was redistributed from 2000, so the goods the money could buy were shrinking while the money grew.
Venezuela's deficit came from an export shock: oil was about ninety per cent of exports, the price collapsed from 2014, and the state held prices and the exchange rate by decree while it kept spending.
Venezuela's physical cash ran out in December 2016, a full year before the hyperinflation is dated as beginning.
All three governments named an external cause (reparations and occupation, sanctions and speculators, sanctions) and acted on that diagnosis with controls and seizures while the financing continued unchanged.
So the shared cause is not a personality, a printing press or a foreign plot: it is a state that must spend more than it can raise, next to a central bank that cannot refuse it.
As the currency dies the exchange rate becomes the price actually used: traders, employers and landlords set prices off the day's dollar rate rather than off yesterday's local price.
The episode transfers wealth in one direction: anyone owing a fixed sum in the currency is released, and anyone holding savings, a pension or a government bond in it is wiped out.
People who can leave, leave: nearly eight million Venezuelans, about a quarter of the country, have gone.
A hyperinflation ends when the government stops needing the central bank to fund it: the regime changes, not the policy.
Germany ended it deliberately, with a new currency in November 1923 and a stop on the government's borrowing from the central bank.
Zimbabwe ended it by abandonment rather than reform: the economy went over to United States dollars, recognised officially in February 2009, which removed the state's ability to create the money it was spending.
Venezuela's ending was not designed: Venezuelans dollarised the economy themselves, the government accepted it afterwards, and the state's reliance on money made by its central bank fell as oil income rose.
All three ended in the same direction: each state came to rely less on money created by its own central bank, and only one reached that by choosing to.
When it ends, it ends fast: the price level stops moving within weeks.
What does not come back is the willingness to hold the state's money: Zimbabwe has relaunched its currency three times since dollarising and all three failed - bond notes in 2016, the RTGS dollar in 2019, and the ZiG in 2024.
The memory outlives the episode and distorts it: German public memory fuses 1923 with the deflation of 1929 to 1933, which was the opposite event.
Ordinary large-scale money creation is not this: after 2008 about three quarters of one central bank's expansion of base money sat as reserves, whereas in 1920s Germany practically all of the new money was spent by the government from the start.
So the test of any monetary expansion is not how large it is but whether the government's spending depends on it.