The filing dropped on a Tuesday, and the first read was predictable. Italy’s largest banking group, Intesa Sanpaolo, cut its reported holdings in BlackRock’s iShares Bitcoin Trust by 93.7%. The headlines wrote themselves: big bank runs from Bitcoin. Then I opened the options section and stopped. A call position tied to 2,496,500 IBIT shares had collapsed to just 18,000. A new put position equivalent to 500,000 IBIT shares had appeared from nowhere. And in that same filing, the bank more than tripled its position in a staked Ethereum ETF while cutting its Solana staking ETF position to seven shares. The ledger remembers what the market forgets. The ledger is not saying sell crypto. It is saying rotate crypto.
Intesa Sanpaolo is not a crypto gambler. It is Italy’s largest banking group by assets, a scale where every new product line moves through compliance, legal, risk and operations before a euro reaches a digital asset. That institutional DNA is why its 13F filings carry weight. It does not trade digital assets for adrenaline. It builds infrastructure.
The bank’s arc into crypto began before the ETF cycle. In July 2024, it underwrote Italy’s first on-chain digital bond on Polygon, a $25.6 million issuance. In January 2025, it made its first direct Bitcoin purchase, acquiring 11 BTC for about $1.03 million. By the time the latest disclosure was filed, Intesa was running a dedicated digital asset desk offering options, futures and spot ETFs. The evolution is not random. It is the sequence of a bank learning the asset class from institutional angles: fixed income on-chain first, token exposure second, derivative distribution third.
Now the numbers. As of June 30, Intesa held 40,723 shares of IBIT. On March 31, it held 646,809. That is a 93.7% reduction. The held-call row fell from an underlying-share amount of 2,496,500 to 18,000, a drop of more than 99%. A put position covering 500,000 IBIT shares appeared for the first time. The staked Ethereum ETF position jumped from 116,200 shares to 349,600 shares, a 201% increase. The Bitwise Solana Staking ETF position dropped from 2,817 shares to seven. Seven. That is not a position; that is a placeholder.
Those numbers will be quoted all week as evidence that a European banking giant is abandoning Bitcoin. That reading is dangerously incomplete. I have audited enough 13F filings to know that the first row is never the whole position. A 13F is an inventory list, not a positioning report. It reports option positions by underlying-share amount, not by delta, not by strike, not by expiration. It does not tell you whether a put was bought or written. It does not tell you whether the call was deep in-the-money or deep out-of-the-money. It treats a covered call strategy and a speculative long call as the same instrument on the same line. If you are going to draw conclusions from a 13F, you have to understand what the form cannot tell you.
Let me outline the forensic problem. In March, the bank had spot IBIT shares plus a call position on nearly 2.5 million additional shares. On the surface, that looks aggressively long Bitcoin. But there are structurally distinct realities hiding behind that surface. The call row could represent a synthetic long position, where the bank bought calls and sold puts to create leveraged upside. It could represent a deep in-the-money call used as a cheap substitute for shares. It could represent a market-making desk holding options inventory because clients were on the other side. The 2.5 million underlying-share amount is real, but the economic exposure depends entirely on parameters the form does not disclose.
Now look at June 30. Spot shares down to 40,723. Call row down to 18,000. Put row up to 500,000. The obvious directional reading is bearish: the bank sold Bitcoin exposure and bought downside protection. But the same data can describe an income desk rotating into a different trade. Suppose the bank had been running a covered call overlay on its IBIT position. It holds shares, writes calls, collects premium. When it sold most of the shares, it repurchased most of those written calls, which is why the call row collapsed. The new put on 500,000 shares could be a protective floor for a structured product packaged for wealthy Italian clients. Or it could be a sold put used to collect premium and re-enter the market at a lower price. The 13F cannot distinguish between a bought put and a written put. It just says put. To call this a bearish banking statement is to ignore the form’s structural ambiguity.
Let me add some rough notional math to show why the share-count comparison is imprecise. The March call position on an underlying amount of 2,496,500 IBIT shares points to a derivatives book that, at the then-prevailing price, would have been in the nine-figure range. The June put position on 500,000 shares points to a meaningfully smaller notional, and the spot line of 40,723 shares is nearly immaterial. That matters. If you are a risk officer, the change is not simply 93.7% off the asset class. It is a shift from a synthetic leverage book to a residual hedging book. That is still a reduction, but it is not the same as a bank dumping its conviction in Bitcoin.
The staked Ethereum ETF line is the real story. Intesa tripled its position in the iShares staked Ethereum product. That is not a small line-item change. It is a 201% increase in a product that has only recently become institutionally quotable. It tells me the bank is not exiting digital assets. It is exiting non-yielding Bitcoin exposure and entering yield-generating Ethereum exposure. This is the signature of a treasury desk, not a retail momentum trade.
Power lies in the code, not the community. The code that powers Ethereum staking produces rewards. The code that powers Bitcoin does not. For Intesa, that is not a philosophical debate. It is accounting. An ETF holding staked ETH can generate a yield stream that flows through the same ledger mechanisms as a dividend. The reward is visible, reportable and usable inside a bank’s net interest margin calculation. Bitcoin’s upside is price-only, and price-only is hard to book for a bank treasury. It consumes risk capital without generating income. In an environment where risk-adjusted returns are under constant boardroom pressure, an asset producing roughly three to four percent in staking rewards is closer to a bond proxy than a meme.
The wrapper matters as much as the token. A European bank cannot casually run validators, manage withdrawal credentials and monitor slashing risk inside a regulated balance sheet. An ETF wrapper delegates all of that operational weight to the issuer. The bank’s accounting department can reconcile a staked ETH position with an ordinary asset register. That is the difference between a trade and an allocation. Intesa already understood this because it ran its own on-chain bond issuance on Polygon; it knows where the seams are.
The Solana staking ETF position is equally informative. Intesa had 2,817 shares, then cut to seven. This is what happens when a bank stress-tests a new asset class and finds the infrastructure lacking. Solana’s staking rewards are nominally higher, but the institutional plumbing is less mature. Liquid staking derivatives are fragmented. The ETF wrapper may not provide the same operational cleanliness. The bank left a placeholder. It did not conclude Solana is worthless. It concluded Solana is not yet a treasury instrument.
The broader US flows complicate the bank-hates-Bitcoin narrative even further. Spot Bitcoin ETFs saw a record monthly net outflow of about $4.5 billion in June. Then July reversed: $172.4 million in net inflows. August has already added another $170 million. IBIT remains the dominant vehicle in the sector with nearly $61 billion in total inflows since launch. This is not a market-wide exodus. If it were, every fund would be bleeding and IBIT would not be the deposit point for hundreds of millions of dollars. The US flow data says the base of Bitcoin ETF holders is still adding, or at least stabilizing. Intesa is not representative of the ETF complex. It is one bank making a portfolio allocation decision.
Meanwhile, BSCN reported last week that BlackRock clients sold on the order of $60 million in IBIT while buying more than $20 million in the firm’s spot Ethereum ETF. That is the same trade, in miniature, that Intesa just printed in its 13F. The rotation from Bitcoin ETF exposure to Ethereum ETF exposure is not a single idiosyncratic decision. It is a pattern.
Here is the angle the headlines will skip: the 94% reduction in IBIT shares may not even represent a 94% reduction in Intesa’s economic exposure to Bitcoin. This is the least understood part of 13F analysis. The shares line is the only line reported with full accuracy. The options lines are reported with their underlying-share amount, but real economic exposure is delta-adjusted. A call option on one million underlying shares can have a delta of zero if the strike is astronomically high, or a delta of one million if the option is deeply in the money. The form does neither. The 2,496,500 call share amount in March may have represented minimal delta if it was a collection of low-delta calls. Conversely, the 500,000 put share amount in June may represent a bullish position if those puts were written, because a written put is a bullish trade. The narrative built on the headline reduction depends on a linear reading of non-linear instruments.
Think about what a bank of this size actually does with options. It does not load up on naked calls because a compliance officer will block that trade. It packages risk. It sells structured notes to high-net-worth clients, then hedges those notes in the listed options market. A note holder gets Bitcoin exposure with a floor at, say, 80% of spot. The bank buys the put as the hedge. Meanwhile it writes calls to finance the floor. The entire package books as options transactions in the 13F. When client demand shifts from Bitcoin-linked notes to Ethereum staking notes, the numbers move exactly the way Intesa’s filing moved. The call row shrinks. The put row appears. The spot inventory shrinks. The yield-bearing Ethereum line grows. The market reads this as conviction; it is actually a distribution pipeline.
From my time on the exchange side, I have seen this pattern in every institutional cycle. The balance sheet is not a mind. It is a routing system. A bank does not have to choose between promising Bitcoin upside and collecting Ethereum yield. It can sell the upside story to clients and keep the cash flow for itself. The 13F shows the residue of that process. It shows which product the clients stopped buying and which product the clients started buying. It does not show the bank’s CFO whispering about Bitcoin. It shows the bank’s product committee noticing that one shelf is generating fees and the other is consuming hedging costs.
There is also a regulatory layer. European banks are not free to hold raw crypto assets the way a US hedge fund can. The ECB and Italian supervisory authorities treat staking as something closer to a financial service than to an anonymous portfolio allocation. An ETF wrapper creates a clean audit trail. Validator risk, slashing risk, smart contract risk and key management are delegated to the fund issuer. That is not a small detail. It is the difference between a trade and an allocation.
Let me address the remaining 40,723 IBIT shares directly. Some analysts will say the bank kept shares, so it still likes Bitcoin. That is equally wrong. For a bank with a balance sheet over a trillion euros, 40,000 IBIT shares is spare change. It is the kind of residual position left behind when a desk closes a larger structured trade but does not bother to liquidate the final tranche, or keeps a sliver to monitor the product’s mechanics. The put covering 500,000 shares is one-hundred times larger than the spot position. The signal is in the ratios, not the absolute share counts.
The market’s instinct is to frame every large bank position as a directional vote. Institutional flows do not work that way. They are structured, hedged, packaged and rebundled. The only reason Intesa’s IBIT position declined by 93.7% is that the number was quoted. That number is the least useful data point in the filing. The useful data point is the 233,400-share increase in staked Ethereum. In a bull market, everyone is looking for confirmation that Bitcoin is alive. The bank is looking for assets that pay them not to wait.
The next filing matters more than this one. Watch whether the staked Ethereum ETF position compounds. Watch whether the 500,000-share IBIT put survives into the third quarter. If the put vanishes and the call row expands, Intesa was packaging downside protection and has repositioned for a different client demand profile. If the put remains and the staked ETH position keeps climbing, Intesa is converting its digital asset desk into a yield machine. The ledger remembers what the market forgets. The market is about to spend a week misreading Intesa’s ledger. The actual question is not whether Italy’s biggest bank believes in Bitcoin. It is whether a bank can book yield from Ethereum’s code in a way its auditors find undeniable. That is the trade behind this trade.