COMEX Silver withdrawals raise questions as physical demand builds [Video]

A bank run is a familiar danger in a fractional-reserve banking system. In December 1930, as the Great Depression deepened, the privately owned Bank of United States in New York collapsed after a run by depositors. Despite its name, it was not a government institution. It served many immigrant families and small businesses.

When rescue-merger negotiations failed and confidence evaporated, depositors rushed to withdraw their money. New York state banking authorities closed the bank on Dec. 11, 1930, leaving more than 400,000 depositors affected. By deposit size, it was the largest U.S. bank failure to that point.

The problem was inherent to fractional-reserve banking. Banks keep only a portion of deposits in reserve while lending and investing the rest. The system functions until too many people want their money back at the same time.

Money Metals Midweek Memo host Mike Maharrey said the silver futures market carries a somewhat similar risk. Futures contracts promise delivery of silver at a specified price and date, but most traders settle or roll their contracts rather than take physical metal. As a result, paper claims far exceed the amount of metal immediately available for delivery.

The exact paper-to-silver ratio is unclear. Maharrey said estimates of 100-to-1 or even 250-to-1 circulate, but he could not find verifiable data supporting them. What can be measured is the amount of registered silver in COMEX vaults relative to paper claims. 

As of mid-September, there were slightly more than five paper ounces of silver for every ounce of registered physical silver, a roughly 5-to-1 ratio.

If a large enough share of contract holders demanded their metal, the system could face significant delivery stress.

More than 7 million ounces leave COMEX vaults

That possibility is worth watching after a notable recent increase in physical silver withdrawals from COMEX vaults.

CME data showed 6,168 September silver delivery notices through Sept. 18, representing 30.84 million ounces. In futures-market terminology, however, “delivery” does not necessarily mean silver physically leaves a warehouse. Usually, delivery means the transfer of a warehouse warrant to a new owner.

The more notable figure was the amount of metal that actually departed the COMEX system. About 7.1 million ounces, or roughly 223 metric tons, left COMEX vaults in the seven days from Sept. 10 through Sept. 17. That represented about 2.1% of total COMEX silver inventory.

Registered silver, meaning metal with an active warrant and available for immediate futures delivery, increased by 1.4 million ounces to 97.3 million ounces during that period. Eligible silver, which meets COMEX standards but lacks an active delivery warrant, fell by about 8.6 million ounces to 232.8 million ounces.

Total inventory consequently dropped from 337.2 million ounces to 330.1 million ounces. The difference shows that approximately 7.1 million ounces physically left COMEX warehouses.

CME reports do not establish whether the withdrawn metal was the same silver associated with September delivery notices. They only show that the metal is no longer in the vault system.

The outflow was roughly 25% larger than the 5.75 million-ounce drawdown between Oct. 3 and Oct. 9, 2025, at the beginning of the October silver squeeze. It was about half the 13.96 million-ounce drawdown recorded between Oct. 9 and Oct. 16, when that squeeze peaked.

Unlike the October episode, though, registered inventories increased last week rather than falling. The earlier squeeze drained nearly 14.2 million ounces from the registered category. The latest data therefore signals strong physical withdrawals, but not the same immediate pressure on deliverable COMEX supplies.

A tight physical market remains the bigger story

One week of outflows does not prove an imminent silver shortage. Metal could flow back into COMEX vaults next week. But sustained withdrawals could become significant quickly and potentially set the stage for another squeeze.

The silver market has already experienced two meaningful squeezes during the past 12 months. The October squeeze pushed silver above $50 an ounce, while a second squeeze sent prices to $120 an ounce in January before a correction. The market was eased in part by shifting metal between London and New York, but Maharrey argued that the underlying supply-and-demand problem has not disappeared.

Silver is on track for its sixth straight annual supply deficit, meaning demand is exceeding newly mined and recycled supply. By the end of this year, the cumulative deficit is expected to approach 800 million ounces, roughly equivalent to one year of global mine production.

Industrial users, jewelry makers, and other consumers must meet shortfalls by drawing on above-ground stockpiles. Before the recent string of deficits, above-ground stocks rose by 243 million ounces from 2010 through 2020. Maharrey said the market has seen a net stock rundown of about 473 million ounces during the last 15 years.

The world is not about to run out of silver, but tighter physical availability creates upward pressure on prices. A continuing pattern of COMEX vault withdrawals could be one factor that lights the fuse for another squeeze.

Rising rates do not automatically mean lower Gold

Maharrey also pushed back against the conventional view that rising interest rates are necessarily bearish for gold and silver. With the 10-year Treasury yield near 5%, its highest range since 2007, some investors have sold precious metals on the expectation that the Federal Reserve will keep tightening.

Higher yields can create an opportunity cost for non-yielding assets such as gold and silver. But the more important measure is the real interest rate: the nominal yield adjusted for price inflation.

A 10-year Treasury yielding 5% while CPI runs at 3.5% has a real yield of just 1.5%. If a bond yields 3% while CPI runs at 5%, the real yield is negative 2%, meaning an investor loses purchasing power in real terms.

Maharrey cited financial writer and Fed watcher Adam Sharp, who pointed to the 1970s as an example. Gold rose nearly 2,329% during that decade even as interest rates climbed. In 1974, three-month Treasury bills yielded 7.4%, but CPI inflation peaked at 12%, producing a real rate near negative 4.6%.

From August 1976 through January 1980, gold rose more than sevenfold while interest rates were climbing toward nearly 20% under Paul Volcker’s inflation fight. Gold also advanced through much of the 2000s bull market as rates both rose and fell.

The lesson is that rate direction alone does not determine gold’s path. When central banks raise rates because inflation is problematic, inflation itself can support demand for hard assets. Maharrey said money-supply growth of roughly 5% annually points to continuing monetary inflation that will eventually appear in consumer and asset prices.

The Fed’s debt trap

Sharp and Maharrey agree that the Fed faces a fundamental Catch-22. It needs higher rates to restrain inflation, but higher rates worsen the government’s debt burden, increase deficits, and threaten an already fragile economy.

Sharp argued that the Fed cannot raise rates to the 10% or higher levels that might be needed to decisively defeat inflation without severely damaging a debt-bloated economy. Maharrey said today’s debt load makes it impossible for Fed Chair Kevin Warsh to operate as another Paul Volcker.

His central takeaway was that inflation remains built into the system. Even the Fed’s stated 2% CPI target means a dollar loses purchasing power every year, amounting to roughly 10% over five years and 20% over 10 years before compounding.

For that reason, Maharrey argued that investors should not sell gold or silver simply because the Fed may hike rates again. He encouraged listeners to consider sound money as a long-term way to protect purchasing power.