Advertiser disclosure
Gold $4,363.10 +27.2% on the year Silver $64.50 +66.2% on the year

11 Sept 2026 live spot

Data tracker · Money supply and gold

Money supply and gold

Money supply and gold have both grown since April 1968, though not together. M2 stood at $23.22 trillion at the July 2026 reading with gold at $4,026.60, so the whole money stock would buy 5.77 billion ounces against a median of 9.33 billion across 700 months. Growth in the two correlates at 0.09 over 12-month windows. Figures come from the Federal Reserve and the LBMA.

Monthly readings, no live figure on this page 700 paired observations since Apr 1968

Ounces of gold the M2 stock would buy

5.77bn

At the July 2026 close, with M2 at $23.22 trillion and gold at $4,026.60.

The dates on those numbers

M2 is published monthly and our series ends at the July 2026 reading, so the figure above is 1 month old rather than current. Our gold series runs 1 month further, to August 2026 at $4,562.75, which is less than a month old and sits 13.3% above the close every paired measure here uses. That difference alone would move the headline reading. For a gold price from this week rather than from a monthly close, our gold price today page carries one.

Where that sits on the 58-year record

5.77
2.27Jan 1980 dearest 9.33Median 19.68Mar 2001 cheapest

Latest readings

Current market snapshot

Both series at the same monthly close, the measure that joins them, and the three tests that decide whether joining them means anything at all. Each card says what the number is and which window it was measured over.

M2 money stock

$23.22T

+5.4% since Jul 2025 close

The Jul 2026 reading of M2 seasonally adjusted, published monthly by the Federal Reserve and the newest month in our series.

Gold, monthly close

$4,026.60

+22.1% since Jul 2025 close

The LBMA fix for the same month as the M2 reading beside it, so neither side of any comparison on this page is dated differently from the other.

Ounces of gold M2 would buy

5.77bn

-13.6% since Jul 2025 close

The whole M2 stock divided by the gold price at the Jul 2026 close, against a median of 9.33bn across the record. A flat line would mean gold was keeping pace with the money supply.

Correlation, 12-month changes

0.09

688 overlapping windows

M2 growth against gold's return over the same twelve months. A reading this close to zero means one number tells you almost nothing about the other.

Same test using inflation

0.35

687 overlapping windows

Measured CPI inflation against the same gold returns, which reads higher than the money supply test does and is the argument most people are reaching for when they talk about money printing.

Months gold was dearer against M2

94 of 700

the most recent was Jun 2026

Gold has been dearer against the money supply than it is now in 94 of the 700 months on record, drawn from 13 separate episodes, so this is a percentile rather than a single event.

Sources: Federal Reserve Board, M2 money stock (SA) and LBMA Gold Price (PM London auction, USD/oz). Correlations run on overlapping 12-month changes in both series. Data verified as of 2026-07-31

Read this before the charts

What the money printing argument claims about gold

The argument runs that dollars are created faster than gold is mined, so each dollar buys less gold as time passes. Stated that loosely, nothing could disprove it. The testable version is narrower, and it says the gold price should rise roughly in step with M2, which is the Federal Reserve's measure of the cash, deposits and money market balances the public actually holds.

Testing it means dividing the whole M2 stock by the gold price, which gives the billions of ounces the money supply would buy in a given month. The argument predicts a flat line. Across 700 months from April 1968 the line has instead run from 2.27 billion ounces in January 1980 to 19.68 billion in March 2001, a range of 9 times.

Gold has still outrun M2 over the full record, compounding at 8.3% a year against 6.7% and finishing 2.4 times ahead. That total-return figure is the one the industry quotes. We earn a commission when a reader opens a gold IRA through this site, which is why the timing evidence below gets published at the same size as it.

M2 since April 1968

43x

6.7% a year across 700 months

Gold over the same months

103x

8.3% a year, price only

Correlation of the two, 12-month changes

0.09

688 overlapping windows

The headline series

Gold measured against the money supply since 1968

The whole M2 stock divided by the gold price at every monthly close, which gives the billions of ounces the money supply would buy. It ran from 2.27 billion in January 1980 to 19.68 billion in March 2001 and stood at 5.77 billion at the July 2026 close, against a median of 9.33 billion.

M2 money stock divided by the gold price, monthly close

Billions of ounces the money supply buys, April 1968 to July 2026

Jul 2026 close

5.77bn

A rising line means the money supply is growing faster than the gold price, while a falling line means gold is outrunning it. The argument predicts a flat one. The dashed rule is the 9.33 billion median, which describes this distribution better than an average does, because the 1979 to 1983 stretch sits far enough from everything else to drag a mean toward itself.

Jan 1980 dearest2.27Mar 2001 cheapest19.68Median 9.33bn 9.33
197019801990200020102020
Billions of troy ounces of gold the whole M2 money stock would buy at each monthly close from April 1968 to July 2026. The series peaks at 19.68 billion in March 2001, bottoms at 2.27 billion in January 1980, and ends at 5.77 billion. The dashed line is the 9.33 billion median.

Sources come from Federal Reserve Board, M2 money stock (SA) and LBMA Gold Price (PM London auction, USD/oz). Data verified as of 2026-07-31

Download the M2 and gold monthly dataset as CSV

Range across the record

2.27 to 19.68

The cheapest reading is 9 times the dearest. Both inputs grew across those years, so a measure travelling that far is not describing a fixed relationship between them.

Months dearer than today

94 of 700

They run from February 1974 to June 2026 and come from 13 separate episodes, so the current reading is a percentile on this measure rather than a single event to compare against.

Against all the gold ever mined

83%

The World Gold Council puts every ounce ever mined at roughly 6.95 billion at the end of 2024. At the July 2026 price the entire US money supply would buy about 83% of that.

Drivers

What moves the money supply and what moves gold

Bank lending, fiscal transfers and deposit flight set the money supply, while central bank buying and jewellery demand set a large share of the gold price, and only real interest rates reach both. That structural gap is why the measured correlation between them sits at 0.09 rather than near either end of the scale.

Each driver, what it does to the money supply, what it does to gold, and which side it lands on
Driver Money supply Gold Reaches
Commercial bank lending Most of M2 is created when a bank makes a loan and credits a deposit account, so private credit demand drives the aggregate more than the Federal Reserve does in an ordinary year. Bank lending has no direct channel into the gold price, and no gold model anybody publishes carries it as an input. M2 only, and it is the largest single driver with no counterpart on the other side.
Federal Reserve asset purchases Buying securities from a non-bank seller creates a deposit and raises M2 directly, which is what happened at scale through 2020 and 2021. Reaches gold through real yields and through expectations rather than through the deposit itself, which is a slower and much noisier path. Both, by different routes and on different timescales.
Fiscal transfers financed by the Federal Reserve Direct payments land in deposit accounts, so M2 grew 40.6% between February 2020 and its March 2022 peak. Gold rose 20.6% over exactly those 25 months, and 8.7% after CPI, so it captured roughly half the nominal money growth. Both, and this is the episode the argument is usually built on.
Real interest rates A positive real yield pulls deposits into money market funds and Treasury bills, which is what shrank M2 by 4.8% between March 2022 and October 2023. Gold pays no coupon, so a higher real yield raises the cost of holding it, which is the one lever that reaches both series in the same direction. Both, which makes it a confounder rather than a confirmation.
Central bank gold buying Foreign official purchases of metal do nothing to the US deposit base, and none of it appears in the H.6 release. Official sector buying has run above 1,000 tonnes a year since 2022 according to the World Gold Council. Gold only, and it is the largest structural buyer with no equivalent on the money side.
Jewellery and mine supply There is no channel at all, because a wedding ring and a mine shaft do not create or destroy a bank deposit. Jewellery is close to half of annual gold demand and mine supply grows roughly 1% to 2% a year, so a large part of the gold price is set by flows that have nothing to do with monetary policy. Gold only, and together these flows set a large part of the price with no reference to monetary policy at all.
Changes to what M2 means The Federal Reserve stopped publishing M3 in March 2006 and replaced the weekly H.6 release with a monthly one in February 2021, so a 58-year M2 chart is not one continuous measurement. The LBMA price is one auction settlement in one metal, and its definition has not changed across our record. M2 only, and it is a reason to treat the long series as an indicator rather than as a constant.

Sources: Federal Reserve H.6 money stock measures, World Gold Council demand trends, and our own monthly series. Price series verified September 2026

The full record

Money supply and gold by decade since 1968

M2 has grown 43 times over since April 1968 while gold has risen 103 times, which is 8.3% a year against 6.7%. The totals agree with the argument and the decades do not, because gold beat money growth in 3 of the 6 decades below and fell outright in 2 of them.

Growth of 100, indexed at the first paired month

M2 and gold from the same start, logarithmic axis

Both lines start at 100 in April 1968 and share one axis, so the vertical distance between them is the whole story. The axis is logarithmic because gold finishes at 10,298 against M2's 4,334. The two lines cross more than once. That is the part the argument does not anticipate, because gold spent the whole of the 1990s falling further behind a money supply that grew in every year of it.

101001,00010,000100,000Starting value 100
197019801990200020102020
GoldM2 money stock
Gold and the M2 money stock both indexed to 100 at the April 1968 monthly close and plotted on one logarithmic axis. Gold finishes at 10298 and M2 at 4334 in July 2026, which is 8.3% a year against 6.7% a year.

Total change across each decade, first paired month to last

Money supply growth against gold, by decade

The money supply grew in every decade on this chart. Gold fell in 2 of them, which is the simplest form of the problem. Almost all of gold's outperformance across 58 years comes from the 1970s and the 2000s, and in both of those decades gold moved closer to the inflation rate than to the money growth rate.

M2 money stockGold
150.0%%1363.3%%112.7%%-39.0%%46.5%%-30.1%%82.4%%283.9%%81.2%%40.4%%50.5%%154.2%%
1970s1980s1990s2000s2010s2020s
Total percentage change in the M2 money stock and in the gold price across each decade from the 1970s to the 2020s, measured from the first to the last paired month inside each one. The 2020s bar runs to July 2026 rather than to a full decade. Hover or tab a decade for both figures and the consumer price change beside them.
Each decade, with the ounces the money supply bought at the start and the end of it
Decade M2 Gold Consumer prices Ounces, start to end
1970s +150.0% +1363.3% +102.9% 16.85 to 2.88
1980s +112.7% -39.0% +61.9% 2.27 to 7.91
1990s +46.5% -30.1% +32.4% 7.63 to 15.98
2000s +82.4% +283.9% +28.4% 16.47 to 7.83
2010s +81.2% +40.4% +18.9% 7.86 to 10.13
2020s +50.5% +154.2% +28.4% 9.74 to 5.77

The 2020s row runs from Jan 2020 to Jul 2026 rather than across a full decade. Sources are the Federal Reserve H.6 release, the LBMA fixes and BLS CPI-U

What we found in the data

Does gold track the money supply

Across 688 overlapping 12-month windows the correlation between M2 growth and gold's return is 0.09, and it changes sign by decade, reading -0.44 through the 1990s. The version of the claim that does survive is narrower: gold's median 12-month gain rises with the money growth band it started in, from +0.7% to +14.3%.

Correlation between money supply growth and gold's return at five horizons, with the number of genuinely separate periods each one contains
Window Correlation Overlapping windows Independent periods How to read it
1 year 0.09 688 58 No usable relationship at this length.
2 years 0.11 676 29 No usable relationship at this length.
3 years 0.16 664 19 No usable relationship at this length.
5 years 0.41 640 11 Reads higher, on only 11 genuinely separate periods.
10 years 0.54 580 5 Reads higher, on only 5 genuinely separate periods.

Computed from 700 paired monthly observations, April 1968 to July 2026. Past performance does not predict future results

Why the 10-year reading is the weakest number in that table

The correlation climbs to 0.54 once the window stretches to 10 years. That reading comes from 580 overlapping windows covering only 5 genuinely separate decades. Two series that both trend upward will correlate across long windows whether or not one drives the other, and 5 observations would not carry a conclusion anywhere else. So we publish the figure and decline to lean on it. The 12-month reading of 0.09 rests on 58 independent years, and it is the one worth arguing about.

Gold's 12-month change, grouped by how fast M2 grew over the same months

What gold did at each rate of money growth

We sorted every 12-month window by the money growth it contained, then measured what gold did across the same months. The medians line up in the order the argument predicts. Gold's median gain was +0.7% when M2 grew under 4% and +14.3% when it grew above 10%, which is the strongest evidence on this page for the claim. One band on the chart breaks that pattern. Across the months when M2 was shrinking gold's median was +7.2%, and it rose in 14 of the 15.

MedianMean
7.2%%8.7%%0.7%%5.8%%5.0%%10.4%%7.6%%13.6%%14.3%%15.5%%
M2 shrinking0 to 4%4 to 7%7 to 10%Over 10%
Gold's 12-month price change grouped by the M2 growth rate over the same 12 months, across five bands covering all 688 overlapping windows between April 1968 to July 2026. Hover or tab a band to read its window count, its episode count and the months it covers.

What the shrinking band is and is not

The 15 months where M2 was smaller than it had been a year earlier all come from 1 run, from Dec 2022 to Feb 2024. That is 1 event rather than 15 readings, and nobody should build an argument on it. It stays on the chart because the money printing case predicts gold struggling through a monetary contraction, and across the only contraction in 700 months of Federal Reserve data gold rose 2.8% instead.

Gold's return delayed behind the money growth it is tested against

Whether gold responds to money growth on a delay

The usual answer to a weak reading is that gold responds late, so we measured the delay rather than assuming it, testing every lag from 0 to 36 months in 6-month steps. The strongest reading is 0.16 at 36 months. The profile falls before it rises, which is not the shape a genuine delayed response makes. Nothing in it reaches the 0.35 that measured CPI inflation delivers against the same gold returns with no lag at all.

0.090.060.050.060.070.110.16
No lag6 months12 months18 months24 months30 months36 months
Correlation between M2 growth over 12 months and gold's 12-month return, measured at lags of 0, 6, 12, 18, 24, 30 and 36 months. The highest reading is 0.16 at 36 months, and every reading sits below the 0.35 that CPI inflation produces with no lag.

The finding that matters

Inflation explains gold better than money supply does

Money growth and consumer prices usually get made as one argument. Across this record the two are only loosely related to each other, correlating at 0.08, which makes them separable. So we ran both against the same gold returns. Across the 687 windows all three series share, measured CPI inflation reads 0.35 against M2 growth's 0.09, and holding inflation constant leaves the money supply with a partial correlation of 0.07.

A reader who buys gold because of money printing is reaching for the inflation argument through a worse proxy, since the Bureau of Labor Statistics already publishes the better one every month. Even 0.35 sits a long way from a relationship anybody should size a retirement position on. Our page on the inflation adjusted gold price takes that half of the question across 58 years of real returns.

None of this says gold was a poor holding, because it compounded at 8.3% a year across the whole record against the money supply's 6.7%. That is a real result and one worth having. The claim the data will not carry is narrower, and it is the timing one: that watching the money supply tells you when to own the metal.

CPI inflation against gold

0.35

M2 growth against gold

0.09

M2 with inflation held constant

0.07

Decades with a negative reading

5 of 6
The same 12-month correlation cut by decade, with the median money growth and the median gold return inside each one
Decade Correlation Windows Median M2 growth Median gold change
1970s -0.12 120 +9.8% +24.1%
1980s 0.03 120 +8.3% -3.3%
1990s -0.44 120 +4.2% -2.4%
2000s -0.31 120 +6.2% +10.9%
2010s -0.03 120 +6.0% +1.5%
2020s -0.01 79 +4.3% +19.5%

5 of the 6 decades carry a negative reading, and the money supply grew in every one of them

What we found in the data

What happened to gold when the money supply tripled from 1980

Between January 1980 and December 1999 the money supply grew 3.1 times over while gold fell 55.6% in nominal terms and 79.5% after consumer prices. A reader buying gold on the money printing argument deserves the strongest case against it, with the dates attached, and this is it.

M2, Jan 1980 to Dec 1999

+212.9%

$1,482.7bn to $4,639.1bn

Gold over the same months

-55.6%

$653.00 to $290.25

Gold after consumer prices

-79.5%

CPI rose +116.4% across the window

Wait for the nominal price back

329 months

Last close below $653.00 was June 2007

Somebody who bought an ounce in January 1980 on exactly the reasoning this page is testing then watched the money supply do precisely what they expected. It grew 3.1 times over across 20 years. The position they bought to guard against that lost more than half its price over the same window. Run it to gold's own low point in March 2001 and the fall reaches 60.5% against M2 growth of 3.4 times, which is 21 years of being right about the mechanism and wrong about the price. Gold did not close below $653.00 again after June 2007, so the nominal wait ran 329 months. The inflation-adjusted wait was longer still, and our inflation adjusted gold price page measures by how much.

Gold's price today if it had tracked M2 exactly from each base month

Why the implied gold price depends on the year you start from

The most common way this argument reaches a price is to take a base year, grow gold by the money supply since then, and publish the answer. That calculation gives $1,293 from December 2000 and $10,226 from January 1980. The spread between them is 8 times, against an actual July 2026 close of $4,026.60. Anybody quoting a single number from this method has chosen a base year, and the base year is doing the work.

Each base month, the prices it starts from, and what it implies for gold at the latest paired close
Base month Gold then M2 then Implied gold now Actual against implied
April 1968 $39.10 $535.7bn $1,695 2.38x
August 1971 $40.65 $685.5bn $1,377 2.92x
January 1980 $653.00 $1,482.7bn $10,226 0.39x
December 1990 $386.20 $3,272.5bn $2,740 1.47x
December 2000 $274.45 $4,927.3bn $1,293 3.11x
December 2008 $869.75 $8,203.0bn $2,462 1.64x
February 2020 $1,609.85 $15,492.8bn $2,413 1.67x

A multiple above 1 means gold already costs more than that base year's money growth implies, while a multiple below 1 means it costs less. Gold sits above 5 of the 7 bases here and below the January 1980 one, which is the base almost nobody in the industry quotes.

The reading we like least, published at full size

At the July 2026 close the money supply buys 5.77 billion ounces against a median of 9.33 billion, so gold sits among the dearest 14% of the 700 months on record on this measure. Only 94 months have ever been dearer, and they come from 13 separate episodes running February 1974 to June 2026. Gold has risen +150.1% since February 2020 while M2 grew +49.9% over the same 6 years, which is what put the reading where it is and is the opposite of the two moving in step.

The two dates behind that reading matter here more than anywhere else on the page. Gold's own August 2026 close of $4,562.75 sits 13.3% above the July 2026 close every paired figure uses, and pairing Jul 2026 money against it gives 5.09 billion ounces rather than 5.77. We keep that number out of the headline because it mixes two dates, and we will not leave it out of the one paragraph where the mismatch moves the answer rather than shading it. None of this is a forecast, and a percentile is not a prediction about next year. It is also the most uncomfortable figure this page produces for a site that earns a commission on gold, which is why it sits here rather than in a footnote.

Timeline

Money supply and gold timeline, with the readings behind each event

Every entry carries the M2 reading and the gold close for that month, the ounces the money supply would have bought, and where that measure stood 12 months later, so the story and the number cannot drift apart.

  1. April 1968

    The record starts, with the dollar still convertible

    Our gold series opens here, 9 years after the M2 series starts, so this is the first month the two can be measured against each other. The dollar was still redeemable in metal at $35 an ounce for foreign central banks, which is worth remembering before reading the first 3 years of this chart as a market price.

    Ounces the money supply buys 13.70bn M2 $535.7bn Gold $39.10 12 months on 13.20bn

  2. August 1971

    Nixon closes the gold window

    When convertibility ended, the money supply would have bought more gold than in any month since our record opened. Everything the money printing argument describes was about to happen from this point on, so it is the fairest possible place to start for anyone who wants to run that argument forward.

    Ounces the money supply buys 16.86bn M2 $685.5bn Gold $40.65 12 months on 11.49bn

  3. January 1980

    The dearest gold has ever been against the money supply

    This is the most expensive gold has been against the money supply in 695 months of data. Anyone who bought that month on the money printing argument waited until June 2007 to see the same nominal price again, and the inflation-adjusted wait ran longer than that.

    Ounces the money supply buys 2.27bn M2 $1,482.7bn Gold $653.00 12 months on 3.17bn

  4. March 2001

    The cheapest gold has ever been against the money supply

    Gold bottomed at the end of a 21-year decline while the money supply had more than tripled since January 1980. Gold lost 60.5% of its price over exactly those years. Both halves of that were happening at the same time and in the same country, which is what makes it the plainest counter-example the record holds.

    Ounces the money supply buys 19.68bn M2 $5,072.2bn Gold $257.70 12 months on 18.25bn

  5. August 2011

    Gold peaks after the first 3 rounds of quantitative easing

    The cycle that began with the Federal Reserve's 2008 asset purchases had by this month more than doubled gold against the money supply. It then spent 4 years giving most of that back while M2 kept growing, which is the half of the quantitative easing story that rarely gets told.

    Ounces the money supply buys 5.26bn M2 $9,532.5bn Gold $1,813.50 12 months on 6.16bn

  6. March 2020

    The CARES Act, and the fastest money growth in the record

    Over the 11 months that followed, M2 grew 26.8% year on year, which is the fastest twelve-month expansion anywhere in 58 years of data. Gold rose 8.3% across those same twelve months to February 2021. This is the episode most readers have in mind when they raise the argument at all.

    Ounces the money supply buys 9.97bn M2 $16,034.4bn Gold $1,608.95 12 months on 11.75bn

  7. March 2022

    M2 peaks at $21.8 trillion

    The expansion stopped here, with gold slightly cheaper against money than it had been when the money creation started. Gold had risen 20.6% over the 25 months from February 2020 while M2 rose 40.6%, so it captured about half the money growth in nominal terms and roughly a fifth of it after inflation.

    Ounces the money supply buys 11.22bn M2 $21,788.1bn Gold $1,942.15 12 months on 10.58bn

  8. October 2023

    The only sustained contraction in the series

    M2 had fallen 4.8% from its March 2022 peak, which is the deepest contraction anywhere in the Federal Reserve's M2 record. Gold rose 2.8% across the shrinkage. If the money supply were the mechanism behind the price, this is the window where gold should have struggled most, and it did the opposite instead.

    Ounces the money supply buys 10.38bn M2 $20,737.5bn Gold $1,996.90 12 months on 7.80bn

  9. July 2026

    The newest month both series cover

    The money supply buys 5.77 billion ounces here against a median of 9.33 billion, which puts gold among the dearest 14% of months on record on this measure. M2 has grown 49.9% since February 2020 while gold has risen 150.1% over the same 6 years, so the two have moved a long way apart in the direction the argument does not usually anticipate.

    Ounces the money supply buys 5.77bn M2 $23,218.0bn Gold $4,026.60 12 months on not yet

Our view

Gold IRA Digest's perspective

Every entry below carries the day it was written. A tracker with one undated opinion sitting under live data is the failure mode this page was built to avoid, so an old note here stays visible as an old note. Nothing on this page passes for current thinking without a date on it.

· latest

The M2 series ends on 31 July and today is 9 September, so the paired figures on this page are a month behind rather than current. Our gold series runs one month further, to 31 August at $4,562.75, which is 13.3% above the July close that every paired measure here uses. One figure moves on that gap rather than being shaded by it. Pairing July money against August gold gives 5.09 billion ounces rather than 5.77, so the newer close makes gold look dearer against the money supply rather than cheaper. We would rather name both dates than quietly pick whichever one flattered the argument.

We reran the correlation after a reader argued that gold responds to money growth on a delay, and the delay does not rescue it. Testing every lag from 0 to 36 months in 6-month steps, the strongest reading was 0.16 at 3 years, which is still weaker than the 0.35 that plain CPI inflation produces at no lag at all. Whatever gold is responding to over these 58 years, the money supply is a worse proxy for it than the inflation number the Bureau of Labor Statistics already publishes.

The Federal Reserve rewrote part of its own M2 history in a recent release, which is worth knowing before anyone cites a long money supply chart. M2SL is seasonally adjusted and those factors get re-estimated, so the March 2001 reading moved 0.05% and October 2023 moved 0.02% without any new information arriving about either month. The revisions are small enough to leave every finding on this page standing, and an LBMA gold fix never moves at all, so a claim resting on both series is only ever as settled as the money half of it.

Written by the Gold IRA Digest Editorial Team and last checked on 9 September 2026. Opinion, labelled as opinion, and never a recommendation to buy or sell anything.

What is scheduled

What to watch next

Windows are months rather than exact dates, because the later release dates are not published yet and inventing one would be the same mistake as inventing a verification date.

  • Monthly, fourth Tuesday Federal Reserve H.6 money stock release The M2 figure at the top of this page, and every paired measure below it Federal Reserve
  • Monthly, mid-month US CPI release The inflation comparison, which reads higher than the money supply test on the same returns BLS
  • Eight times a year FOMC rate decision Real yields, which is the one lever that reaches the money supply and the gold price together Federal Reserve
  • Weekly, Thursday Federal Reserve H.4.1 balance sheet The asset purchase side of money creation, ahead of the monthly M2 print Federal Reserve
  • Sep 2026 Our monthly refresh of both series Ours Every figure on this page, and the six-month gap in the paired reading with it Gold IRA Digest
  • Oct 2026 World Gold Council Q3 demand trends Central bank buying and jewellery demand, which sit on gold's side of the driver table alone World Gold Council
  • Nov 2026 Our rerun of the correlation and band study Ours The correlation figures and the growth bands, recomputed once another two quarters of paired months exist Gold IRA Digest

What this means if you hold metal in an IRA

What money supply growth means for a gold IRA

The money supply is a poor timing signal on this evidence, so it is a poor reason to move a retirement account in a hurry. A salesperson quoting it is quoting a relationship that reads 0.09 over 700 months of data. A gold allocation can still be a reasonable holding for reasons this page does not touch, and gold and equities have spent long stretches moving apart, which our gold vs the S&P 500 page measures directly.

A self-directed IRA holds the metal under section 408(m), which sets a purity standard of 0.995 for bullion bars and names the depository your coins sit in. Those rules do not move when the money supply does. Your outcome turns on the premium over spot the day you buy, the annual fee your custodian charges, and the storage fee the depository bills on top of it. Our page on gold IRA fees carries what each custodian we called charges, with the date we checked it.

A dealer who opens on money printing and will not put a fee schedule in an email has told you which of those two numbers they would rather you looked at.

Common questions

Money supply and gold questions

Does gold go up when the money supply goes up?
No. Across 688 overlapping 12-month windows from April 1968 to July 2026, M2 growth and gold's return over the same 12 months correlate at 0.09. That is close enough to zero that one number tells you almost nothing about the other. The relationship also changes sign by decade, reading -0.44 through the 1990s while M2 grew in every year of it. One version of the claim does survive. Gold's median 12-month gain rises with the money growth band it started in, from +0.7% when M2 grew under 4% to +14.3% when it grew above 10%.
How much did gold rise when the money supply grew 40% after 2020?
Gold rose +20.6% between the February 2020 and March 2022 monthly closes, while M2 grew +40.6% over those same 25 months. Consumer prices rose +11.0% across the window, so gold gained +8.7% in real terms. The fastest 12-month money growth anywhere in the record ran to February 2021 at 26.8%, and gold rose 8.3% over exactly those 12 months. The largest monetary expansion on record produced a single-digit gold return.
What is the gold price if it tracked the money supply?
There is no single answer, because the number depends entirely on which month you start from. Running M2 growth forward from December 2000 implies $1,293 an ounce today, while running it from January 1980 implies $10,226. The actual July 2026 close was $4,026.60. A calculation whose answer moves by a factor of 8 on the choice of base year is a rhetorical device rather than a valuation, and anybody quoting one figure from it has picked the base that suits their argument.
Is gold expensive relative to the money supply right now?
Yes, on this measure and at the July 2026 close. The whole M2 stock would buy 5.77 billion ounces of gold at that close, against a median of 9.33 billion across 700 months, and only 94 months in the entire record have been dearer than that. Those 94 months come from 13 separate episodes rather than one stretch, so the reading is a percentile rather than a comparison to a single event. Gold's own August 2026 close paired against the same money supply gives 5.09 billion ounces, which mixes two dates and is why the headline uses the paired close instead.
Is inflation a better guide to gold than the money supply?
Yes, though neither is strong. Measured CPI inflation correlates with gold's 12-month return at 0.35 across 687 windows, against 0.09 for M2 growth over the same returns, and holding inflation constant leaves M2 with a partial correlation of 0.07. In other words the money supply adds almost nothing to what the Bureau of Labor Statistics already publishes. Our page on the inflation adjusted gold price measures the inflation half of that question against 58 years of real returns.
Does gold respond to money supply growth on a delay?
No. We tested every lag from 0 to 36 months in 6-month steps, and the strongest reading was 0.16 at 36 months, which is still weaker than the 0.35 that plain CPI inflation delivers with no lag at all. The reading falls before it rises, which is not the shape a delayed response produces.
What happened to gold when the money supply tripled after 1980?
Gold fell 55.6% between January 1980 and December 1999, while M2 grew 3.1 times over. Consumer prices rose +116.4% across those 20 years, so gold lost 79.5% of its purchasing power. It last closed below its January 1980 price of $653.00 in June 2007, which is 329 months after that peak.
How often does this page update?
The Federal Reserve publishes M2 monthly in its H.6 release, and we rebuild both series from the published closes, so every date here comes from the data rather than from a clock. No live money supply figure and no live gold quote is wired into this page. The reading at the top is the July 2026 paired close, and it carries that label everywhere it appears.

How this page is built

The headline measure is the whole M2 money stock divided by the gold price, both in US dollars. M2 comes from the Federal Reserve H.6 release seasonally adjusted and gold uses the LBMA London fixes, and both were verified as of 2026-09-09, giving 700 paired observations from April 1968. The two pass through one join that collapses any repeated month on the side it walks and keys the other side by month, so a month printed twice in a source file cannot reach the arithmetic as two months. M2 publishes 2 months behind our gold series, and the join walks M2, so every paired figure here runs to July 2026 while gold's own record runs to August 2026. The page names which is which wherever both appear.

Correlations run on overlapping percentage changes rather than on the levels, because two series that both trend upward correlate on levels whatever they are doing. Overlapping windows are not independent observations. Every correlation therefore carries its window count and its count of genuinely separate periods, and every band in the growth study carries its episode count beside its window count. Real returns use BLS CPI-U and appear only inside the two named episodes, since our real gold price page owns inflation adjustment across the whole record. Every chart renders on the server and works with JavaScript disabled, which is also why the assistants quoting this page can read it.

Sources

The full dataset is available as an M2 and gold monthly CSV, and corrections go to the editorial team and are logged publicly.

Gold IRA Digest is an independent publisher. This page is information rather than investment advice, and nothing on it recommends buying or selling anything. Past performance does not predict future results. Some links on this site earn us a commission, which never changes what we publish about a company or an asset.