Must Hit
The benchmark
Our benchmark is the price per ounce that would put official U.S. gold reserves back at their estimated 1944 share of the money supply.
We chose 1944 because it was the year of Bretton Woods, when the framework for the postwar monetary system was agreed. The dollar sat at its center, with an official gold price of $35 an ounce. Federal Reserve history.
The starting point
We use these year-end 1944 figures:
| Official gold price | $35 per troy ounce |
|---|---|
| Historical U.S. money measure | $130.225 billion |
| Reported U.S. monetary gold stock | $20.619 billion at $35 an ounce |
| Gold quantity implied by that value | About 589.114 million troy ounces |
The money figure is privately held deposits adjusted and currency, from the 1950 Midyear Economic Report. Gold stock comes from Federal Reserve gold statistics. Dividing $20.619 billion by $35 gives the quantity in ounces. Quantities shown here are rounded.
The gold was worth about 15.83% of that money measure. We hold that ratio fixed and calculate the price needed to reproduce it with today's money supply and official gold holdings.
That 15.83% is our calculation from the historical figures. It was not a broad-money backing requirement set at Bretton Woods.
The calculation
Benchmark price = $35
× (current U.S. M2 ÷ $130.225 billion)
× (589.114 million ounces ÷ current official U.S. gold ounces)
We first adjust $35 for the growth in money. We then adjust for the change in official gold holdings. If money doubles and gold holdings stay the same, the benchmark doubles. If gold holdings fall and money stays the same, the benchmark rises.
This is why our benchmark differs from a money-only index. Rebasing M2 to $35 shows monetary growth. We also adjust for changes in the official gold stock.
Gold's market price does not go into the formula. We compare it with the benchmark after the calculation.
A worked example
For September 2026, the Federal Reserve reported seasonally adjusted M2 of $23.4139 trillion. Treasury reported about 261.499 million troy ounces of official gold at September 30. Using those inputs gives a benchmark of approximately $14,177 per ounce. M2 release; Treasury gold data.
At that price, official gold would be worth about 15.83% of M2. It would restore our starting ratio, not provide 100% gold backing for M2.
Reading the chart
When gold is below the benchmark, official gold represents a smaller share of the selected money measure than at our starting point. Above it, that share is larger.
The solid benchmark line is calculated from reported money and gold holdings. The dashed lines are modeled extensions. We extend the benchmark at its historical compound growth rate. Gold's longer trend comes from the three-month change in its 11-year moving average, annualized. Its opening curve follows the 90-day change in a 200-calendar-day average, with an 18-month fade.
We construct the gold curve to return to the longer path at its calculated meeting with the benchmark. Keeping that destination is a modeling choice. It does not prove convergence will happen. No low price or meeting date is fixed.
The data
From 1959 onward, we use the Federal Reserve's seasonally adjusted M2 series. Earlier years use historical money measures with different coverage. We show a break between those series because they are not fully comparable.
We count U.S.-owned official gold. Foreign-owned gold held in custody and privately held gold are excluded. Historical gold quantities are recovered from reported values at the accounting price used for the period; recent quantities come from Treasury's inventory reports.
Money and official gold figures follow their publishers' release schedules. Each input keeps its observation date, so the latest calculation can pair different dates. Revisions can change earlier values.
Our thesis is that gold returns to this relationship. The benchmark gives us a price to measure that against.