Some thoughts on a US gold audit and revaluation.
I discuss past audits of US gold holdings, the accounting associated with revaluation, whether it has any implications for US Treasury purchases or sales, and its potential implications for policy.
There has been no comprehensive audit of US official gold holdings, although to incomplete audits have been conducted, in 1953, and in 1974. A key factor constraining an audit being undertaken is simply the logistics of doing so. It’s an immense undertaking.
Revaluing the US Treasury’s gold holdings would require an Act of Congress, but the accounting between the US Treasury and the Federal Reserve is straightforward.
There is no need for the US Treasury to sell gold, nor is there any requirement for them to maintain a price floor once the gold is revalued.
There are two key questions that need to be answered regarding revaluing the US Treasury’s gold holdings: 1) What would the US Treasury do with these funds? and: 2) How might the US Federal Reserve respond?
Should the US audit its gold holdings?
There have been audits conducted in 1953 and again in 1974. My reading of the historical reports indicates that both audits were partial and were not independently verified, so they cannot really be described as complete. From the perspective of 2025, an audit today would likely comprise US Treasury/Mint officials plus an approved external auditor (e.g. agents from one of the Big Four accounting firms) plus an external advisor from an approved refinery, to confirm the assay tests.
However, it’s important to understand the logistical challenges involved.
To give an example, in the 1953 audit, the US Mint team opened three of the twenty-two sealed compartments at Ft. Knox depository, and counted 88,000 gold bars, with 9,000 of those gold bars weighed. The US Mint then conducted assay tests on 26 gold bars chosen at random, drilling holes from the top and bottom. All were found to be “of a fineness equal to, or in excess of that appearing in the Mint records and stamped on the particular gold bars involved.” To conduct a complete audit of the entirety of the US gold stock, including gold held in Denver, West Point, and the New York Federal Reserve would, simply, be a logistical nightmare. So, after the initial flurry of excitement of opening the vaults, presumably with photo opportunities for the President & Mr Musk, getting down to the real job of auditing the holdings would be a laborious task.
In their report (HERE) the 1953 audit concluded “on the basis of assays, your committee can positively report that the gold represented, according to assay, is at the depository. We have no reason, whatsoever, to believe other than, should all melts be assayed, the results would be the same. We, the undersigned, found the assets verified, to be in full agreement with the assets as indicated by the joint seals affixed to the respective compartments, on January 26, 1953. It is the opinion of this committee that the same agreement would be found should all of the compartments be verified."
The 1974 audit was conducted by the US General Accounting Office and the Department of the Treasury and included agents from the US Customs Service, Bureau of Government Financial Operations, and the Bureau of Mint. As with the 1953 audit, there were no independent auditors and assayers involved in the operation. The operation consisted of opening three compartments and counting 91,404 bars, “or about 21% of the gold stored at the depository.” The committee weighed a selection of the gold bars and took assay samples to verify the gold’s fineness.
Consistent with our view that logistics are a key challenge in auditing, the Committee’s report (HERE) notes “the tremendous quantities of bullion in the Mint depositories make it impossible , for practical reasons, to weigh, count, and assay, and check all values at each settlement.”
I conclude that there is probably a good case to be made for a comprehensive audit, including independent auditors and assayers, but whether the benefits outweigh the costs remains unclear.
What would be the impact on the US Federal Reserve’s balance sheet?
The mechanism by which the US Treasury and Federal Reserve interact with respect to the Treasury’s gold holdings is rather straightforward. The US Treasury holds the gold on its account, but it provides the Federal Reserve with Gold Certificates, with gold currently valued at USD 42.2222/oz, this amounts to a book value of USD 11,037m. And on the liabilities side of the balance sheet, the Federal Reserve makes these funds available to the US Treasury by crediting the US Treasury General Account.
So, to revalue the US Treasury’s gold holdings, it would simply require an Act of Congress adjusting the value, to say USD 2,900/oz, with the Federal Reserve re-valuing its Gold Certificates, increasing its value to USD 758,068m, with the Federal Reserve then crediting the US Treasury’s General Account to the tune of USD 747,031m (i.e. the Gold Certificate value minus its initial value).
Does this mean the Federal Reserve or US Treasury must buy gold to protect its value?
The short answer to this is no.
The longer answer is that the government is not like you and I. As gold is classified by the US government as a commodity, not as a monetary instrument, the accounting rules are unclear and there is no mechanism for them to be required to maintain any mark-to-market (which is why we are where we are) and so if they endure a mark-to-market loss on their holdings, they simply don’t need to realize it. In a way, this is like bonds held on the balance sheet of a central bank. Yields may rise (bond prices fall), but the central bank is under no obligation to mark the bond to market. Where the loss crystalizes is when the bond is redeemed. However, gold can be thought of as a perpetual bond, so there is no maturity date, and no enforcement mechanism to deliver a loss (if prices fall).
So, there is no need for the US Treasury to underwrite the gold market (i.e. talk of a new Gold Exchange Standard) by buying gold at that price. Nor is there any need for them to sell the physical asset, as they can simply mark the commodity to market and draw down the valuation change as a claim.
What would the Treasury do with these funds?
As the funds would be deposited into the Treasury’s General Account at the Federal Reserve, these funds would become available for the Treasury to use as directed by the Executive. Absent a budget incorporating the Trump Administration’s policy agenda, it’s really impossible to say where these funds might be directed. I personally see four possible avenues: i) they seed a US Sovereign Wealth Fund, ii) they fund a Bitcoin Reserve, iii) they provide scope for additional spending, and/or: iv) they get absorbed into the Treasury’s pool of funds, with an offset in reduced bond issuance.
Given the US Government’s wide primary (excluding debt servicing) deficit and elevated Federal public debt held by the public, my preference would be for these funds to be used to offset debt issuance. This would likely lead to a modest decline in US longer-term bond yields and the US dollar, providing a limited boost to US asset prices, while being the least inflationary outcome. Will this happen? Probably not given the political dynamics in Washington DC.
How might the US Federal Reserve respond?
This would depend in large part on what the US Treasury does with the funds. To the extent that the funds are used to offset existing bond issuance, the impact could be benign. However, if the Treasury deploys these funds, they would almost certainly have an expansionary impact on the US money supply. With the US labor market tight, additional fiscal stimulus could become inflationary, leading to a policy response by the Federal Reserve, pushing up interest rates, and the US dollar.
The information in this publication is provided by Redward Associates Private FZCO for informational purposes only. The insights contained in this publication are not investment advice and should not be construed as such, nor as a recommendation for any specific investment product. Our full disclaimer can be found HERE.
Thank you Peter. Very clear.