← Back to home

Money Supply and Inflation: Testing the Monetarist Promise

"Inflation is always and everywhere a monetary phenomenon." Milton Friedman's famous claim has guided central bank policy for decades. Print too much money, get inflation. Simple, right? We test 65 years of data (1960-2025) and find something uncomfortable: the correlation between M2 money supply growth and CPI inflation is -0.06. Essentially zero. The Fed printed $3.5 trillion during QE (2008-2014) with no inflation. Then printed less in 2020-2021 and got 9% inflation. What's going on?

The Monetarist Theory

Milton Friedman's monetarism rests on a simple equation: MV = PQ (the "equation of exchange"):

The monetarist claim: if you hold V and Q constant, then changes in M must cause proportional changes in P. Print twice as much money, prices double. The Federal Reserve was literally founded on this principle in 1913, and Volcker's inflation-fighting campaign (1979-1982) was guided by it.

But what if V doesn't stay constant? What if it collapses? That's the story of the past 17 years.

Test 1: The Long-Run Relationship (1960-2025)

Let's start with the most basic test: does M2 growth predict inflation over the long run? We'll plot year-over-year M2 growth against year-over-year CPI inflation for every month from 1960 to 2025.

M2 Growth vs. CPI Inflation (1960-2025)

Each dot represents one month. Correlation: -0.06 (essentially no relationship). If monetarism were correct, we'd see a strong positive correlation (upward-sloping cloud of points).

Result: No Correlation

Correlation coefficient: -0.06. A perfect correlation is 1.0. Random noise is 0.0. We're seeing basically random noise. Over 65 years and 791 months of data, there is no meaningful relationship between M2 growth and inflation.

This doesn't mean money supply is irrelevant—it means the relationship is more complicated than "print money → get inflation." We need to dig deeper.

Test 2: The QE Puzzle (2008-2019)

The most dramatic test of monetarism happened after the 2008 financial crisis. The Federal Reserve launched three rounds of Quantitative Easing (QE), expanding its balance sheet from $900 billion to $4.5 trillion. M2 money supply grew by 70% from 2008 to 2019. Monetarism predicts massive inflation. What actually happened?

The QE Era: Money Printing Without Inflation

M2 money supply (blue) exploded 70% from 2008-2019. CPI inflation (orange) stayed below 2.5% the entire time, averaging 1.5%. Where did the inflation go?

The QE Period Stats

Metric 2008-2019
M2 Growth (Total) +70%
Average Annual M2 Growth 6.3%
Average CPI Inflation 1.5%
Peak Inflation 2.5% (2011)

Result: Monetarism failed spectacularly. The Fed printed $3.5 trillion in new money and inflation stayed dead. This should have been impossible according to the theory.

Test 3: The 2020-2023 Inflation Spike

Then came COVID. In 2020-2021, M2 grew by 40% (the fastest two-year increase since World War II). This time, inflation surged to 9% by mid-2022—the highest since 1981. Monetarism vindicated?

COVID Era: Money Printing WITH Inflation

M2 grew 40% in 2020-2021 (less than the QE era!), but this time inflation hit 9%. What changed?

The COVID Period Stats

Metric 2020-2023
M2 Growth (Total) +40%
Average Annual M2 Growth 13.5%
Average CPI Inflation 4.8%
Peak Inflation 9.1% (2022)

The Puzzle: M2 grew less than during QE (40% vs 70%), but inflation was much higher (9% vs 1.5%). Something other than money supply must explain the difference.

Test 4: The Missing Variable - Velocity

Remember MV = PQ? We've been focusing on M, but what about V (velocity of money)? Velocity measures how many times each dollar is spent per year. If velocity falls, you can print lots of money without causing inflation. If velocity rises, even moderate money printing causes high inflation.

Velocity of Money (M2V): The Key Variable

Velocity of M2 peaked in 1997 at 2.2 (each dollar was spent 2.2 times per year). It crashed to 1.1 by 2020—a 50% collapse. This is why QE didn't cause inflation: the money sat in reserves, it didn't circulate.

The Velocity Story

1997-2008: Velocity was stable around 2.0. M2 growth translated predictably into GDP growth + moderate inflation.

2008-2019 (QE Era): Velocity collapsed from 2.0 to 1.4 (-30%). The Fed printed money, but banks held it as reserves. Households deleveraged instead of spending. Corporations bought back stock instead of investing. The money didn't circulate, so no inflation.

2020-2021 (COVID): Velocity hit an all-time low of 1.1, but then stimulus checks went directly to consumers who immediately spent them (unlike QE, which went to banks). Supply chains broke, so spending hit fewer goods. Result: demand spike + supply collapse = inflation.

2022-2025: Velocity stabilized around 1.2. M2 growth slowed (even contracted in 2023). Inflation fell back to 3-4%.

What Actually Drives Inflation?

Based on 65 years of data, here's what matters:

  1. Money supply × Velocity, not just money supply alone. You need both M and V to rise for sustained inflation.
  2. Who gets the money matters: QE went to banks (low velocity). COVID stimulus went to consumers (high velocity).
  3. Supply shocks matter: 2021-2022 had shipping disruptions, semiconductor shortages, energy price spikes, labor shortages—all independent of money supply.
  4. Expectations matter: Once inflation expectations become unanchored (2022), wage-price spirals can sustain themselves even as money growth slows.

The Real Lesson

Monetarism isn't "wrong"—the equation MV = PQ is an accounting identity, it must be true by definition. But it's incomplete. Friedman assumed V was stable or predictable. It's not. V collapsed 50% from 1997 to 2020, then spiked in 2021-2022. Without understanding velocity, money supply tells you almost nothing about future inflation.

Implication for investors: Don't trade on M2 growth alone. Watch velocity (though it's only reported quarterly with a lag). Watch who is getting the money. Watch supply chains. Watch fiscal policy (stimulus checks vs. bank reserves). Inflation is a multi-variable problem, not a single-variable one.

Comparing to Other Inflation Measures

We've used CPI (Consumer Price Index) so far, but the Fed targets PCE (Personal Consumption Expenditures). Does using PCE change the story?

M2 Growth vs. Multiple Inflation Measures

M2 growth (blue), CPI inflation (orange), PCE inflation (green), Core CPI (red). All inflation measures track each other closely, but none correlate well with M2 growth. The spike in 2020-2022 is visible across all measures.

Inflation Measure Correlations with M2 Growth

Inflation Measure Correlation with M2 Growth
CPI (Headline) -0.06
PCE (Fed's Target) -0.08
Core CPI (Ex Food/Energy) -0.03

Result: Doesn't matter which inflation measure you use. None of them correlate with M2 growth. The monetarist relationship is broken across all common metrics.

Key Takeaways

Data & Methodology

Data Sources:

Calculations: