In the past few days, so many things happened in both areas I have been covering extensively for a very long time: AI frauds and Silver. While the two topics might appear very distant from each other, perhaps the only thing in common is how silver is increasingly used to produce high-performance semiconductors, today I would like to show you how, in reality, a broken and out-of-control system of rapacious investment banks stands between the two.
Before diving in, allow me to make one thing very clear: most of the bankers I know are good people, often very intelligent, stuck in highly paid, tedious jobs, within hyperbureaucratic institutions, that give them little purpose in life besides making a high salary, while their talent could be of much better use for society elsewhere. Then there is the very small minority, but very important one of the “Mavericks”. In this category, you find the greedy and ruthless ones who would cut any possible corner to achieve their goals, both in terms of career and personal gain. These people are very dangerous, but rest assured, they don’t want to end up in jail if anything goes wrong. As a result, the way they tend to operate is to sit at the very edge of what’s legal and what’s not; morality isn’t even in the back of their minds. However, when you operate in this way, at some point you will breach the law just a little bit without getting caught. Once this happens, you will have strong confidence that breaching the law a second time won’t put you in trouble. One last aspect of these specific types of bankers is the “I know it all” attitude, while the truth is, any person out there doesn’t know what he doesn’t know. This type of attitude tends to make people blind to any potential mistake or bad consequence of their actions. Putting all of this together, these are constantly the type of people that break the system, take very bad and selfish decisions, ultimately putting not only their bank and career at risk, but even the taxpayers. Some examples? From Nick Leeson to Jerome Kerviel, from John Meriwether to Bruno Iksil, to the fraudsters Bernie Madoff and Bill Hwang.
Now it should be easier to understand all I am going to present next about the AI bubble and Silver.
Have you noticed how, even after the great circular financing scheme among hyperscalers has been publicly exposed, banks didn’t withdraw their support from any of those companies, even though it is becoming clearer and clearer how a good chunk of what we still call AI industry is a giant financial fraud? The answer to this is straightforward: by keeping the AI circular financing scheme fed, banks are making a ton of revenue with, so far, little amount of direct exposure. Furthermore, if the regulator says nothing is illegal, then you are golden, right?
Why have I said banks have so far little amount of direct exposure to the AI fraud? Because, similar to the mortgage securitization business that started in 2004 and imploded spectacularly in 2007, ushering in the great financial crisis of 2008, today the banks still find plenty of their fund-managing clients (Private Credit, Pension Funds, Insurances and so on) willing to buy all the toxic AI paper they are able to churn out. The most glaring example that just happened is the absurd investors’ demand that banks still have to distribute Oracle’s debt (“Oracle shares gain as $50 billion raise eases data-center funding fears“), despite the fact that not only is it undeniably unsustainable for the company’s present operations, but it is also being raised on the premise that OpenAI will be able to pay the $300 billion purchase commitments to Oracle, something that is technically impossible as I explained in “AI IS BECOMING ORACLE’S LITTLE BIGHORN“. Why are banks marketing, supporting, and distributing debt on behalf of an issuer that they know has a very high chance of damaging other clients? Because they make a ton of risk-free fee revenues in doing so, and with their buyers being “professional investors” that generally shield them from any potential repercussions. Hilariously, we have a very fresh example of this, considering some investors are trying to sue Oracle and its brokers for distributing Oracle bonds late last year, bonds where they are now losing money, claiming the true state of Oracle’s debt burden was being misrepresented: “Oracle sued by bondholders over losses tied to AI buildout.”
Do you realize how totally ridiculous and broken this is? Trust me, I can spend hours on the topic, provide plenty of examples, but I believe I made my point very clear here. If you feel you already watched a similar movie between 2005 and 2008, in 2000, or even before, I confirm you are not mistaken. As is typical of Wall Street, the same script is recycled with different actors and settings since this is the only way they know how to make a lot of money quickly. Lending to the real economy, to real businesses, and waiting to collect their interest is a slow process that doesn’t help much if you have to constantly goose your share price higher and higher and distribute handsome payouts to shareholders without expanding your balance sheet disproportionately. Let’s move on to silver now.
While in the AI circular financing bubble, fraud, whatever you want to call it, banks do have the incentive to inflate the whole structure as much as possible because the larger it grows, the higher the amount of fees they make in the process. In the case of silver and other precious metals, the opposite is true. Why? Because banks aren’t commodity traders in a strict sense, hence trading the real metal, shipping it from suppliers to customers from one part of the world to another. Banks are financial intermediaries in the commodities market that, in simple terms, make money by selling and transacting in paper contracts written on top of the real commodity. Consequently, it should not come as a surprise that in precious metals and other commodities, the paper market is multiple times larger than the real-world physical market. In the case of silver, more than 200 times larger. In technical terms, banks do make money in this whole space by being constantly short volatility and ultimately suppressing the natural price discovery of a commodity driven by real demand and supply dynamics. Through paper contracts and re-hypothecation dynamics, banks effectively inflate the supply of a commodity disproportionately in the financial market, and this is very profitable because, effectively, they are “selling” one piece of the real stuff multiple times. When does this all break? When those who bought a financial contract with the option to ask for the delivery of the metal want to receive the physical one, not a cash payment.
In this chart below, courtesy of Eric Young, you can see how this whole system worked for many years between the Comex and the LBMA till buyers from China and India started to ask for more and more of silver to be physically settled and delivered.
Because of the chronic shortage of supply compared to the constantly growing demand, as I described in my previous analyses like “WHY SILVER CAN HIT $100 BY MARCH“, “THE RISK OF A SILVER CHAOS” and “THE FINAL ACT: HOW THE LBMA COULD UNRAVEL“, industrial buyers from China and India started to drain the vaults of the Comex, the LBMA and the Shanghai Exchanges at a faster and faster speed to get all the metal they needed.
Intuitively, the lower the amount of silver underneath the huge pile of paper contracts, the more unstable the whole paper contracts house of cards lying upon it becomes. Furthermore, because derivative prices are tied to the price of their underlying, in this case silver, in a structure where the whole system is short of the metal while paper remains abundant, a rise in prices intuitively works against those who wrote those paper contracts. Why? Because when a system based on cash settlements shifts to physical settlement, banks that before considered their exposure “hedged” will find themselves in a difficult situation of seeing those hedges stop being effective.
For instance, when a mining company goes to a bank to hedge the price risk of its future mining, all it expects from a bank is to receive or pay the cash settlement that covers the gap between the price agreed and the price at maturity of the contract. That mining company already has customers who will be receiving the metal, hence the banks won’t receive physical delivery. How did banks “hedge” this long exposure from the miners’ contracts? By shorting silver in the OTC (when they found an opposite demand) or in the futures market, like the Comex. Here is the catch: when in the futures market the ones on the long side are clients that look to purchase physical metal because they won’t receive enough supply from the miner, hence not purely hedging a price risk anymore, the banks who are short those contracts will find themselves in the situation of delivering the metal while on the other leg of their exposure they will instead receive a cash settlement. How do banks find the metal not to default on those short contracts? They either buy or lease it from counterparts that have extra physical inventory in their vaults. When those counterparts become unwilling to let go of their inventory to the banks, is when everything breaks, and effectively, banks become “naked short”.
What happened last Friday and yesterday is the ultimate consequence and desperate attempt of banks desperately trying to source silver. Physical silver ETFs like the SLV are a sure source of physical, but if banks just buy SLV shares and redeem them, that will create additional buying pressure on the underlying asset, compounding their problem and ultimately making their “naked short” positions more expensive to close. Knowing how brokers completely missed the structural shift in the silver market, and instead of quickly getting out of their mess months ago, assuming it was only a temporary issue, they doubled down on their “naked shorts.” The ultimate proof of the acute scarcity of silver physical bars is the widening premium between the Comex/LBMA and the Shanghai exchanges. Why? Because if the banks could easily source the physical metal, like in the past, they would have immediately exploited that arbitrage. Ultimately, the last resort for the banks was raiding the bars held in the physical ETFs, and that’s exactly what they did last Friday.
How could this be done? The answer can be easily found in the latest SLV ETF 10-Q filing:
Here is the whole explanation of how they successfully exploited the arbitrage last Friday as I posted on X during the weekend:
Comex futures price settlement at the Comex is based on a VWAP between 13:24 to 13:25 EST.
LBMA price settlement instead happens at 12:00 UK time.
Most of Silver OTC contracts settle using the LBMA reference and many OTC contracts expire into month end.
On the 30th Jan, LBMA silver price benchmark settled at $103.19, while the Comex Benchmark a few hours later, settled at $78.29.
It is an open secret now how many banks and brokers were underwater on their silver positions, gold, and other precious metals, especially after the rally in January.
What’s even more remarkable is how precious metals crashed on Friday in isolation; stocks, bonds, and other commodities were totally unaffected. Anyone who understands any basic of macro and markets knows how this is logically wrong.
Let’s add one more piece to the puzzle now: on Friday, the OI at the Comex decreased by 8k contracts at the end of the day. Assuming, for simplicity, the price differential between the LBMA and the Comex as a reference. That means banks were able to extract ~$1 bn gain from their shorts, pushing the Comex price through the floor after the LBMA settlement.
Furthermore, the SLV ETF continued trading after the LBMA benchmark settlement creating almost a 20% discount to NAV because of that. Here is the other trick the banks pulled. If you have to settle a lot of physical contracts at the LBMA, but you don’t have the metal, because of such a discount to NAV, Authorised Participant banks could buy SLV shares in the open market from panic-selling investors, tender the shares to claim bars at $103.19, and make a killing in the process.
Not surprisingly, SLV shares count increased by ~51m shares from Thursday to Friday, according to iShares. Because of the NAV discount, banks extracted up to $1.5 bn of profits exploiting this ETF, assuming they bought up all those new shares, arbitrage, and then turned around to claim silver bars at a much higher price for contract settlement purposes.
This exact same scheme was replicated on Monday, even though to a far lesser extent, considering how an almost 40% drop in prices exhausted the selling and enticed new demand to come in.
How many shares did the banks manage to redeem in-kind? I am afraid we will only know the answer to this question in about 6 months. However, as you can see from the latest SLV 10-Q, banks already started to redeem SLV shares in-kind for physical silver in large quantities already in Q2 and Q3 of 2025, even before the LBMA crisis of October and before the sharp rally in silver prices began after Thanksgiving.
I hope this article has been helpful to open everyone’s eyes and help in understanding the perverse incentive Wall Street has to keep a broken financial system in place for their own profit till it ultimately becomes unsustainable and falls apart, requiring extreme interventions, often illegal, like Friday’s silver crash or even a public bailout, to make sure the global financial system does not implode.