The Identity Fisher Actually Wrote Down
Irving Fisher published the equation of exchange in 1911 in The Purchasing Power of Money. It reads MV = PY: the money stock times how often it changes hands equals the price level times real output. Every dollar spent is a dollar received, so the two sides measure the same national spending from opposite ends. The identity holds at any moment, for any country, with no assumptions attached.
The formula becomes a forecasting tool once you express it in growth rates. Money growth plus velocity change, minus real output growth, leaves the growth rate of prices — that is the calculator's default mode. Enter m, v, and y as percents per year and it returns the exact inflation rate together with the popular linear shortcut, so you can see the size of the approximation error you would be taking.
Fisher also gave finance the interest-rate relation (1 + i) = (1 + r)(1 + π) that links nominal and real rates through expected inflation. If your question is how price history changed the value of a dollar across years rather than how money mechanics drive prices, an inflation calculator built on index data is the better fit; this tool stays on the money side of the family.
The Quantity Theory Assumption That Makes It a Forecast
An identity by itself predicts nothing. The quantity theory of money adds two assumptions: velocity moves slowly because payment habits and technology change slowly, and real output is set by capacity, labor, and productivity rather than by money. Under those assumptions inflation ≈ m − y, and money growth becomes the driver of prices.
Milton Friedman compressed it into the line that inflation is 'always and everywhere a monetary phenomenon.' The 1970s fit the story: money growing near 9% a year, velocity rising about 2%, and real growth near 3% give 7.94% inflation on the exact identity — the same high-single-digit CPI zone that dominated 1974 through 1979. You can measure how those prints compounded with the CPI inflation calculator.
The assumptions break at both extremes. In 2020, US velocity collapsed by more than 20% within months, far faster than any payment-habits story allows, and the money-output link bent with it. In hyperinflations, output and velocity both move along with money, so the identity still balances even while the theory's causal story needs more care.
Exact Arithmetic vs the m + v − y Shortcut
Textbooks quote π ≈ m + v − y because the cross terms stay small when the rates are small. The default inputs (7, −1, 2.5) show the size of the error: the shortcut says 3.50% while the exact multiplicative identity returns 3.35%, a 0.15 point gap. At 1970s-scale inputs (9, 2, 3) the gap narrows to 0.06 points — 8.00% versus 7.94%.
The error depends on the products of the rates involved, not their size alone. Negative velocity growth partially cancels money growth inside the cross terms, which is why the 1970s case lands closer than you might expect. Once any input reaches double digits, trust the multiplicative form; compounding is the same force that drives any compound growth calculator.
The identity also splits spending growth cleanly: nominal GDP growth is (1 + m)(1 + v) − 1, exactly 5.93% at the defaults. Real growth absorbs 2.5 of those points and prices carry 3.35. That decomposition — how much new spending buys more stuff versus costlier stuff — sits at the core of every central-bank inflation postmortem.
Choosing the Money Stock: M1, M2, or Currency
The identity is only as good as the M you feed it. M1 (currency plus checking deposits) tracked spending well for decades until 2020, when stimulus deposits blew it up more than 40% in a year without matching price moves. M2 adds savings and small time deposits and has been the monetarist standard since Friedman — US M2 sits near $21 trillion, which is the calculator's default stock.
Match the stock to the price index you care about. M2 against the GDP deflator keeps the units consistent economy-wide, while currency in circulation against CPI works better for emerging markets where cash dominates transactions. Whatever stock you pick, compute its actual velocity as nominal GDP divided by that stock — the level mode performs exactly this division for you.
The 2020-2023 round trip is the cautionary tale: M2 grew about 25% in a single year, then contracted roughly 2% a year through 2023, the first sustained shrinkage since the 1930s. Balances parked in savings earning almost nothing (an APY calculator shows how little that paid) sat completely still — and idle money exerts no price pressure until it starts moving again.
Velocity: The Term That Moves When You Stop Watching
Velocity is nominal GDP divided by the money stock — how many times per year an average dollar funds a purchase. US M2 velocity ran near 2.0 in the late 1990s, about 1.4 on the eve of the pandemic, bottomed near 1.1 through 2020-21, and climbed back toward 1.4 by 2024. It is a measured ratio, not a constant, and it swings by more than most forecasts budget for.
The 2020 case explains a year of missing inflation: money grew 25%, but velocity fell about 20% while real output dropped 2.5%. Run those through the calculator and prices rise just 2.56% — the money created and the money spent were different quantities. Those swollen cash balances then lost buying power when velocity normalized and the delayed inflation arrived in 2021-22.
In 2021 the sign flipped: nominal spending grew about 11% against 6% real growth, leaving roughly 5% for the deflator, and in 2022 the split ran near 9% nominal, 2% real, and close to 7% inflation. Velocity moving with money instead of against it doubled the price pressure — the exact identity captures this interaction, while the shortcut only approximates it.
Level Mode: From Money Stock to Price Index
Switch the mode select and the calculator works in levels: money stock times velocity is nominal GDP, and dividing by real output gives the price level as a GDP deflator. The defaults — $21,000B times 1.4 turns — produce $29,400B of nominal spending and a deflator of 133.64 against $22,000B of base-year output. That 133.64 is the implied price index: output costs 33.64% more than in the base year.
The deflator moves inversely with real output when money is held fixed. Grow output to $23,000B while holding M and V constant and the deflator falls to 127.83, a 4.35% drop in the price level. Supply-side growth is deflationary in the identity's world — more output funded by the same spending means cheaper units of that output.
For long horizons, compound the rate rather than multiplying a yearly change by the number of years. At 7% money growth with velocity flat and real growth at 2.5%, prices rise 4.39% a year and double in about 16 years; shut real growth off and the doubling takes 10.2 years. It is the same doubling math that runs any compound interest calculator, applied to prices instead of balances.
Stress Tests: Depression, Disinflation, Hyperinflation
The Great Depression is the identity's darkest validation. From 1929 to 1933 the US money stock fell about 30% as bank failures destroyed deposits, velocity fell roughly 20%, and real output fell 26%. The exact identity returns a 24.3% price-level drop; the actual CPI decline through 1933 was about 25%. Falling money, falling velocity, and falling output compose a deflation machine.
Japan since the 1990s runs the opposite stress test. Quantitative easing expanded the monetary base enormously, but broad money grew only around 3% a year while velocity drifted down about 1% against 0.5% real growth — which works out to 1.46% inflation. The identity explains why two decades of base-money expansion never produced sustained 2% inflation: the created money sat as reserves and idle balances instead of turning into spending.
Hyperinflation is the quantity theory's home turf. When money doubles monthly, output and velocity changes are rounding errors and π ≈ m almost exactly. Zimbabwe in 2008 and Weimar Germany in 1923 printed at that pace, and black-market exchange rates tracked the exploding money stock almost tick for tick — the same derived-rate arithmetic you can run with a cross exchange rate calculator.
Using the Identity as a Forecasting Discipline
The calculator works backward too. To hold inflation at 2% with velocity drifting down 1% a year and real growth at 2.5%, money must grow 5.61% — that is (1.02 × 1.025 / 0.99) − 1. Central banks that targeted money growth explicitly, as the Fed briefly did from 1979 to 1982, were running exactly this arithmetic in reverse every quarter.
Treat the output as a consistency check rather than a prophecy. Velocity is endogenous — it responds to the same rates and expectations you are trying to forecast — and money shocks hit prices with a one-to-two-year lag. The identity tells you which combination of m, v, and y is compatible with an inflation target, and it exposes which assumption you are really arguing about when you dispute a forecast.
Money-side and rate-side analysis pair naturally. This tool prices the quantity channel, while the Fisher effect calculator handles the interest-rate channel — what nominal rate delivers a given real return once expected inflation is set. Run both on the same scenario and the money story and the rate story have to agree with each other.