Velocity of Money: Why It Fell, and What It Actually Measures

Velocity of money is nominal GDP divided by the money supply. It measures how many times a unit of money turns over buying finished goods and services in a year. US M2 velocity reads 1.412 for Q2 2026, on the Federal Reserve’s own M2V series. We hold neither half of that ratio as FRED builds it. We hold broad money for 185 economies — a wider aggregate than the M2 sitting in FRED’s denominator — and no nominal GDP at all. That gap is why this page explains the mechanism instead of rebuilding a chart FRED already owns.

What velocity of money means

Velocity answers one question. For every dollar of money supply in the economy, how many dollars of spending did it support in a year? Divide nominal GDP by the money stock and that’s the number. US M2 velocity is 1.412 for Q2 2026 (FRED M2V, released 2026-07-30). Each dollar of M2 supported about $1.41 of nominal spending. A higher number means more transactions per dollar in circulation. A lower one means money is sitting still relative to what the economy produces.

The idea traces to Irving Fisher’s 1911 equation of exchange, MV=PQ: money supply times velocity of circulation equals the price level times real output. Milton Friedman’s monetarism leaned on the same term. Friedman’s version of the quantity theory assumed velocity was stable enough to treat as fixed. That assumption is load-bearing. Money neutrality rests on it — the claim that more money moves only nominal variables like the price level in the long run, and leaves real output alone. We trace the identity’s full history, and where the stable-velocity assumption breaks, at quantity theory of money.

Velocity is an identity, not a behavior

Nobody counts how many times a dollar changes hands. Velocity is calculated, not observed: nominal GDP divided by the money supply. That makes it an identity, not an independent behavior. “Velocity fell” and “M2 rose faster than nominal GDP” are the same sentence, not two separate facts.

Financial commentary often treats velocity as if it moves on its own, driven by consumer psychology or bank caution. It then reads a falling number as new information about spending. That’s not wrong, exactly. It’s incomplete. The psychology has to show up as GDP growing more slowly than the money stock before it can show up in velocity at all. That ratio is the entire content of the identity. Nothing else is in there.

The classical dichotomy is the assumption underneath: that the real and nominal sides of the economy stay cleanly separable. Velocity is where you find out whether they did.

Our own series shows the denominator’s half of this — with one caveat that matters. US broad money grew 6.6% a year on average over the last decade (10-year CAGR, through 2025-12). It grew 6.0% in the year to that date. Both figures are the IMF’s harmonised broad-money aggregate, not the Fed’s M2, so they do not slot into FRED’s M2V ratio. Whether velocity rose or fell over any stretch of that decade depends on whether nominal GDP grew faster or slower than the Fed’s own M2. We hold neither of those. That is why there is no US velocity chart of our own on this page, and why there won’t be one until we hold a GDP series.

Why velocity fell after 2008

The 2008 financial crisis pushed banks toward holding reserves rather than lending them out. The Federal Reserve’s response amplified that. The Fed’s own balance sheet grew roughly fivefold between 2008 and 2014, reaching about $4.5T (Federal Reserve H.4.1 release). That expansion sat mostly in bank reserves at the Fed, not in loans that would have funded new spending. Reserves aren’t part of M2. Money parked as a reserve balance doesn’t buy anything. The broad money that was growing moved through the economy more cautiously than before the crisis — deposits held, not spent or lent forward. Nominal GDP grew too slowly, relative to the money stock, to hold velocity where it had been. We cover the same 2008-2014 balance-sheet expansion, and why it didn’t translate into matching inflation, at money supply vs inflation.

Why velocity fell again after 2020

A second, sharper drop followed the 2020 pandemic shock. This time the mechanism ran through households directly, not bank reserves. Lockdowns cut spending options. Stimulus payments and expanded unemployment benefits added income at the same time. The US personal saving rate spiked to roughly 33% in April 2020 (Bureau of Economic Analysis) — money received and set aside, not spent.

US broad money grew 23.5% year over year by Q2 2020, on our own IMF harmonised series, while nominal GDP contracted sharply in the same quarter. A money stock growing fast, dividing into an economy producing less: velocity had to fall, by construction. It fell further and faster than in 2008. Money sitting in a savings or checking account, not turning over in transactions, is exactly what a falling velocity ratio describes. Same mechanism, twice, twelve years apart, at a larger scale the second time. The transmission mechanism from money growth to prices runs partly through velocity, and a falling velocity is exactly what slows that transmission down.

Is high velocity of money good?

Depends entirely on why it’s high. Rising velocity in a recovering economy usually means renewed confidence. People and firms spend rather than sit on cash, and nominal GDP grows faster than the money stock. That reading treats a velocity rise as healthy. A second, very different mechanism pushes the same number up for the opposite reason. In a currency losing value fast, holding cash for even a day is a loss. Everyone spends the moment they are paid. Velocity spikes as a symptom of collapse, not confidence — the classic pattern inside a hyperinflation.

Neither FRED’s own M2V page nor Investopedia’s velocity entry draws that distinction where a reader asking “is high velocity good” would find it (both checked 2026-08). If you know a public page that does, tell us and we’ll link it. We hold no velocity series for any hyperinflation episode ourselves. The mechanism is well documented independently of our numbers, and we cover the money-supply side of those episodes at hyperinflation. The honest answer: the number alone doesn’t say. The direction of nominal GDP relative to the money stock does, and so does why that direction is happening.

What we hold, what we don’t, and where the rest lives

We hold broad money for 185 economies, at whatever frequency each central bank reports it, with 157 carrying a full 10-year growth history. We hold no nominal GDP series for any country. So we hold no velocity figure of our own, for any country — not the United States, not any economy in our broader dataset.

Two separate gaps sit between us and a velocity number, and it is worth being exact about both. The first is the numerator: nominal GDP, which we do not have. The second is the denominator: our aggregate is the IMF’s harmonised broad money, wider than the Fed’s M2, so even with a GDP series our ratio would not be the M2V anyone is quoting. That is a hole in our sourcing, not a design choice.

A genuine cross-country velocity table is still the natural next step for this page. Building it needs nominal GDP, or the World Bank’s FM.LBL.BMNY.GD.ZS series — broad money as a share of GDP, whose reciprocal is velocity. Neither is in our data yet. Until one is, this page states the mechanism rather than estimating a number we can’t verify.

For the US figure itself, FRED’s M2V is the primary source. It publishes the result directly, so we link to it rather than rebuild it. The link between velocity’s fall and the lag from money growth to prices is traced at quantity theory of money and money supply vs inflation. How we source and compute our own broad money is set out at methodology. Every term above is defined at our glossary.

FAQ