What is the difference between M1, M2, and M3?

The short answer

Think of monetary aggregates as a series of expanding circles. The smallest circle, M1, contains money you can spend immediately — paper bills, coins, and balances in checking accounts that you can write checks against or use with debit cards. M2 expands the circle to include money that requires one easy step to spend, like savings accounts and money market mutual funds. M3, where it exists, expands further to include money locked into time deposits over $100,000.

The Fed redefined M1 significantly in May 2020, moving savings deposits from M2 to M1 because they had become functionally equivalent to checking accounts in modern banking. This reform created an artificial level shift in M1 data — historical comparisons require care.

Why discontinue M3? The Federal Reserve concluded in 2006 that M3 added little forecasting power beyond M2 while requiring substantial data collection. The decision was controversial — some argued it concealed information about institutional money creation — but the Fed has stuck with the change.

→ New to monetary aggregates? Financial education hub

What the data shows

FRED data on M1 (M1SL) shows the May 2020 redefinition created an apparent quintupling of M1, which jumped from approximately $4 trillion to over $16 trillion essentially overnight — an accounting reclassification, not actual money creation.

Key figures (FRED and ECB, 1980-2024) :

• M1 jump from May 2020 redefinition : ~$4tn → $16tn+ overnight
• M1 level (April 2024) : ~$18tn
• M2 growth 1980→2010 : $850bn → $8.2tn
• M2 peak (February 2022) : $21.7tn
• M2 contraction 2022-2023 : ~$1tn (first sustained decline since 1949)
• Eurozone M3 1999→early 2024 : €4.6tn → €15.5tn
• Eurozone M3 peak growth 2020-2021 : ~12% YoY
• M2 velocity (FRED M2V) : ~2.2 (1997) → ~1.1 (2022)

The exception worth noting: the velocity of M2 has fallen ~50% over 25 years. This shift typically translates to each dollar of M2 supporting roughly half the economic activity it did in the late 1990s — a dynamic that significantly complicates the interpretation of any aggregate as a forecasting tool.

→ Dataset: US M2 dataset

Why it happens — the macro mechanism

The progression from M1 to M3 reflects three layers of monetary economics that have evolved over decades.

Liquidity gradient. The aggregates are ordered by ease of use as means of payment. M1 captures money that functions immediately as transaction balances. M2 captures money that requires minor friction to spend. M3 captures money that is essentially financial wealth in liquid form. The boundaries shift as financial technology evolves — the 2020 M1 redefinition reflected that savings accounts had become as accessible as checking accounts in practice. See money supply explained.

Money creation channels. M1 is created mainly by central bank operations and currency demand. M2 is created mainly by commercial bank lending — when a bank issues a loan, it creates a deposit in M2. M3 was created by institutional money market activity. Tracking each layer revealed which channels were active. Linked to how the Fed creates money.

Forecasting utility. Different aggregates have different relationships with inflation, output, and asset prices. The Fed’s 2006 decision to discontinue M3 reflected econometric work suggesting M3 added little to M2-based forecasts. The eurozone disagreed and kept M3 as their primary aggregate. This methodological divergence reflects different theoretical commitments about what monetary aggregates measure.

The 1979-1982 Volcker experiment with monetary targeting revealed the practical difficulty of using these aggregates for policy. Targeting M1 or M2 produced extreme interest rate volatility without clean inflation outcomes. This experience accelerated the global shift toward interest rate targeting, with monetary aggregates relegated to diagnostic rather than operational status.

The aggregates are concentric circles of liquidity — the larger the circle, the looser the connection to spending decisions.

→ Framework: Monetary regimes

What it means for different economic actors

Bond investors watch M2 growth as a partial inflation signal, though the relationship is unstable. Periods of rapid M2 contraction often coincide with credit stress and falling bond yields.

Equity investors have observed correlations between M2 growth and equity returns, particularly during liquidity-driven market cycles like 2009-2021. The relationship is weaker during fundamentals-driven cycles.

Eurozone observers need to track M3 specifically because it is the ECB’s primary aggregate. M3 growth signals can differ meaningfully from M2 signals, particularly when institutional money market activity diverges from retail deposits.

A common error is treating M1 series as continuous across the May 2020 redefinition. The level shift requires either splicing the series or using M2 for long-run analysis. Researchers often use the savings deposits at commercial banks series (FRED SAVINGS) to reconstruct pre-2020 M1 equivalents.

Practical observation

Go deeper

Related questions

• What is the money supply?
• Why did central banks abandon monetary targeting?
• How does money supply affect stock returns?
• Does printing money cause inflation?

Frequently asked questions

Why was savings deposits moved from M2 to M1 in 2020?

The Fed amended Regulation D in March 2020, eliminating the six-per-month limit on certain transfers from savings accounts. This made savings deposits functionally equivalent to checking accounts for transaction purposes. Statistical practice followed: starting in May 2020, savings deposits were reclassified into M1 to reflect their actual liquidity characteristics. The change was methodologically defensible but created data discontinuity that researchers must handle carefully when analyzing monetary trends.

Does the eurozone M3 give different signals than US M2?

Often yes. M3 includes institutional money market funds and large time deposits that M2 excludes. During periods of institutional money market activity — like the 2020-2022 ON RRP surge in the US — eurozone M3 may show patterns invisible in US M2. The ECB’s monetary pillar references M3 directly, while US monetary policy has moved away from aggregate references. Comparative analysis requires either using equivalent definitions or focusing on growth rates rather than levels.

Why does monetary aggregate data still matter if central banks no longer target it?

Three reasons. First, aggregates remain diagnostic indicators of liquidity conditions and credit dynamics. Second, financial stability analysts watch them for signals of credit booms or contractions. Third, the long-run quantity theory relationship — money growth eventually translating to inflation — has not been refuted, only loosened. The 2020-2022 episode where massive M2 expansion preceded inflation by 12-18 months reminded analysts that aggregates retain some forecasting content for extreme regime shifts even if they fail at high frequency.

Last updated — 28 April 2026