Quantity Theory of Money

The equation, line by line

Irving Fisher's "equation of exchange" is deceptively simple:

M × V = P × Q

Read each variable carefully:

• M — money supply. Usually M1 (cash + checking) or M2 (plus savings, money-market funds). Choosing the aggregate is half the battle.
• V — velocity. The average number of times each unit of M is spent on final goods and services in a year. Calculated, not observed: V ≡ (PQ) / M. So this side of the equation has one unknown.
• P — price level. The GDP deflator or CPI.
• Q — real output. Real GDP.

The product PQ is nominal GDP. As written, the equation is a tautology — it must be true by construction. The theory appears when we add an empirical claim: V is roughly constant in the short run because spending habits and payment technology change slowly, and Q is roughly constant in the short run because the economy is at full employment. With both held fixed, percentage changes in M map one-for-one onto percentage changes in P.

Differentiating logarithmically gives the version macro students memorize:

%ΔM + %ΔV = %ΔP + %ΔQ
inflation = money growth + velocity drift − real growth

If money grows 5%, real growth is 2%, and velocity drift is zero, inflation should be 3%. That is the textbook prescription, and the Friedman "k-percent rule" — grow M at a steady 3–5% per year — falls straight out of it.

Worked example: Fisher's classroom problem

Suppose an economy has:

• M2 = $1,000 billion
• Real GDP (Q) = $2,000 billion (in base-year dollars)
• GDP deflator (P) = 1.00

Then V = (P × Q) / M = (1.00 × 2,000) / 1,000 = 2.0. Each dollar of M2 is spent twice on final output per year — exactly the average historical US value.

Now the central bank doubles the money stock to $2,000 billion. If V holds at 2.0 and Q holds at $2,000 billion (full-employment ceiling), the new identity reads:

$2,000B × 2.0 = P × $2,000B
P = 2.0

The price level doubles. That is the quantity theory's signature prediction: monetary expansion at full employment is fully inflationary.

Now relax the assumption. Suppose Q can rise — there are unemployed workers and idle capacity. With Q jumping to $2,400B and V drifting up to 2.1 from a stronger economy:

$2,000B × 2.1 = P × $2,400B
P ≈ 1.75

Inflation is 75% rather than 100%, the rest of the monetary expansion is absorbed by real growth. This is the Keynesian objection in algebraic form: when V and Q are not pinned, the simple link from M to P weakens.

Quantity theory vs Keynesian liquidity preference

	Quantity theory (Friedman)	Liquidity preference (Keynes)
Core equation	MV = PQ	M_d = L(Y, i)
Velocity	Stable, predictable	Endogenous, can collapse
Money demand	Function of Y mainly	Function of Y and interest rate i
At zero rates	Theory holds, expect inflation	Money demand goes infinite (liquidity trap)
Policy prescription	k-percent rule, stable M growth	Active counter-cyclical fiscal + monetary
Inflation cause	Always monetary, in the long run	Demand-pull or cost-push, often non-monetary
Central bank target	Money aggregate	Interest rate or unemployment
Empirical strength	Long-run, cross-country	Short-run, recessions, ZLB

The empirical record

The strongest evidence for the quantity theory is cross-country and long-run. Stanley Fischer (1981) and George McCandless and Warren Weber (1995) documented that across 110 countries over 30 years, the correlation between average money growth and average inflation is roughly 0.9. The relationship is essentially one-for-one — every extra percentage point of money growth shows up as a percentage point of inflation.

Hyperinflations are quantity-theoretic events without exception:

• Weimar Germany, 1923. Money base growth ~30,000%/month, inflation ~30,000%/month.
• Hungary, 1946. The largest hyperinflation in history, 41.9 quadrillion %; near-perfect tracking of money issuance.
• Zimbabwe, 2008. 89.7 sextillion % at peak. Money printer ran 24/7.
• Venezuela, 2018. Cumulative inflation 130,060% in a year, monetary base up similarly.

The weakest evidence is short-run and inside liquidity traps. From 2008 to 2014, the US monetary base went from about $0.85T to $4.0T, an increase of 370%. Cumulative CPI inflation over the same window was about 11%. The textbook quantity theory predicts ~370% inflation; reality delivered one-thirtieth that. The reason is the velocity collapse documented in the liquidity-trap literature: M2 V fell from 2.0 to 1.4, almost exactly canceling the M expansion when looking at MV.

Counterarguments and modern critiques

Three serious challenges to the simple QTM:

1. The endogeneity critique (post-Keynesian). Money is not exogenous — central banks accommodate the demand for money created by bank lending, rather than the other way around. Causation runs from PQ to M, not from M to PQ. Charles Goodhart's eponymous law captures the practical consequence: any monetary aggregate the central bank targets becomes useless as a target.
2. The velocity-instability critique. If V is not stable, the M→P arrow doesn't hold even directionally. Financial innovation has made V unpredictable since 1985.
3. The fiscal theory of the price level (Cochrane, Sims). The price level is determined by fiscal solvency expectations, not money quantity. Even at constant M, inflation can erupt if the public expects future fiscal deficits to be monetized.

Variants and refinements

• Cambridge cash-balance equation. M = k × P × Y, where k = 1/V is the fraction of nominal income held as money. Mathematically equivalent to Fisher but psychologically focused on demand.
• Friedman's k-percent rule. Set %ΔM = (target inflation) + (long-run real growth). For 2% inflation and 2% real growth, the central bank should grow M at 4% per year, every year, forever.
• Nominal GDP targeting. Scott Sumner's market monetarism repackages QTM as targeting MV (= nominal GDP) directly, sidestepping the velocity-instability problem.
• New Monetarist economics (Lagos–Wright). Rebuilds QTM with explicit search frictions, deriving V endogenously instead of assuming it.
• MMT. Inverts QTM. Money is not exogenous; the binding constraint on inflation is real-resource availability, not money quantity. Friedman would call this self-contradiction; MMT economists see it as the natural endpoint of the endogeneity critique.

Common pitfalls

• Treating MV = PQ as predictive without checking V. The equation is an identity. The prediction requires extra assumptions, and those assumptions fail at the ZLB and during financial-innovation phases.
• Picking the wrong M. M0 ≠ M1 ≠ M2. The base went up fivefold in 2008–2014 but M2 only doubled. Predictions that quoted "money supply" without specifying which one were essentially noise.
• Confusing levels and growth rates. "Doubling M doubles P" is the levels claim. "5% money growth → 5% inflation" is the rates claim. Both follow from the equation but require different empirical conditions.
• Forgetting the long-run qualifier. Friedman always said money matters "in the long run." Critics who quote the short-run record to refute monetarism are attacking a strawman; defenders who quote it to confirm monetarism overreach in the other direction.
• Ignoring expectations. The 1970s lesson: anticipated money growth feeds straight into wage and price-setting. Surprise money growth has different effects from announced money growth — a refinement Lucas formalized in the 1970s.