The number everyone will quote is 93.7%. Intesa Sanpaolo cut its reported IBIT position from 646,809 shares to 40,723 between March 31 and June 30. Italy's largest banking group, trimming Bitcoin exposure by nearly 94%. Institutional conviction, fading. Another headline for the bears.
The headline is correct arithmetic and wrong analysis.
I've read 13F filings long enough to distrust them. They are net positions captured at a single timestamp — a photograph of a moving object. They don't reveal intent, hedging structure, or the difference between a directional exit and a risk remodel. The details buried beneath the share count tell a different story entirely.
Check the options grid. The held-call row collapsed from 2,496,500 underlying shares to 18,000. A 99.3% reduction. In the same disclosure, a new put position covering 500,000 IBIT shares materialized. The naive reading: Intesa flipped bearish on Bitcoin. The trader's reading: this is the fingerprint of structured repositioning, not a conviction exit.
Before I explain why, let me establish who this buyer is.
The Bank That Was Never a Tourist
Intesa Sanpaolo is not a newcomer to digital assets. It is the kind of institution that does homework before it touches a chain. In July 2024, it used the Polygon network to underwrite Italy's first on-chain digital bond — $25.6 million. That wasn't a marketing experiment. It was a settlement infrastructure test, the kind serious banks run before committing real balance sheet.
By late 2024, the bank had opened a dedicated digital asset desk, offering options, futures, and spot ETFs linked to digital assets. In January 2025, it made its first direct Bitcoin purchase: 11 BTC at roughly $1.03 million. Small. Deliberate. The scale of a pilot program, not a conviction bet.
Every position Intesa has taken in this market carries the signature of a risk laboratory. Positions respond to internal tolerance frameworks, hedging policy, and regulatory conversations with the Bank of Italy — not to Twitter sentiment. That matters because it changes how you interpret the data.
So when the second-quarter 13F landed, my question wasn't "is the bank bearish on Bitcoin?" It was "what is the new risk architecture telling us?"
The answer is more interesting than the narrative.
Core: What the Options Grid Actually Reveals
Let me walk through the mechanics, because this is where the forensic details matter.
A 13F reports holdings at quarter-end. For options, it reports the underlying share amount — not the premium paid, not the strike, not the expiration date. That opacity is deliberate. Regulators want transparency; market participants don't. The result is a filing that tells you enough to be dangerous, and not enough to be certain.
Here is what we know, stripped to the numbers.
The share position: 646,809 to 40,723. Down 93.7%.
The call position: 2,496,500 to 18,000 underlying shares. Down 99.3%.
The put position: zero to 500,000 underlying shares. A new entry.
On its face, this is a massive reduction in total IBIT exposure. But break it down like a risk book and a different pattern emerges.
First, note the asymmetry in the original disclosure. As of March 31, the call position covered 2,496,500 shares — roughly 3.8 times the actual share position. You do not hold calls on 2.5 million shares while owning only 647,000 unless you are running an options strategy that involves writing calls. This is the signature of a covered call program, or a collar structure — using premium from call sales to subsidize downside protection or boost yield.
I've seen this exact architecture on institutional books before. When a bank offers options exposure to clients, the principal book accumulates covered calls as a byproduct of that flow. It is a yield-enhancement vehicle, not a directional thesis.
Now look at the June 30 structure. Shares cut to 40,723. Calls cut to 18,000. Puts added on 500,000 shares.
Here is the contradiction most analysts missed. If this were a bearish exit, you would expect the bank to sell shares and buy puts covering roughly the remaining position. That's textbook risk reduction. But a protective put on 500,000 shares against a 40,723-share position is not a hedge. It is twelve times the share position. Puts covering 500,000 shares while holding 40,000 is not protection. It is either a directional overlay, or the remnant of closing a much larger book.
I ran similar analyses during the Terra collapse in 2022, when I coded scripts to track on-chain inflows into exchanges rather than follow the panic. The same lesson applied: the visible position at the end of a trade tells you nothing about the sequence of steps that got there. What matters is the structure.
The structure here suggests three possibilities.
One: Intesa's options desk carried inventory from client flow. The 99% call collapse aligns with reducing a client-facing book, not expressing a macro view. Banks that write options for high-net-worth clients carry principal positions as a byproduct. The second-quarter unwind could simply reflect client demand shifting.
Two: The bank closed a yield-enhancement strategy. Covered call writing caps upside. If internal models — or conversations with regulators — signaled a potential Bitcoin breakout, the rational move is to unwind short calls and buy downside protection instead. You preserve upside participation, cap the gap risk on the downside, and maintain regulatory composure.
Three: The put position hedges a separate, undisclosed exposure. Banks hold crypto exposure across subsidiaries, custody relationships, and lending books. A 500,000-share IBIT put could offset risk elsewhere on the balance sheet.
All three are more plausible than the "bank is bearish on Bitcoin" narrative.
Now look at what happened to the rest of the book.
The Ethereum Staking Rotation
While the IBIT position contracted, the iShares Staked Ethereum Trust position tripled — from 116,200 shares to 349,600 shares. A 200% increase.
Read that alongside the Bitwise Solana Staking ETF position, which collapsed from 2,817 shares to seven. Seven. Not a rounding error. A statement.
This is the part of the filing that matters most for forward-looking analysis: Intesa is consolidating its digital asset exposure into one instrument — staked Ethereum — while pruning everything else.
Why staked ETH?
First, yield. The staked Ethereum trust generates cash flow. In a bear market, institutional allocators do not want price exposure alone. They want yield to cushion drawdowns. ETH staking offers roughly 2.5 to 3.5% annually in current conditions. That is not trivial for a bank balancing a risk-weighted asset framework.
Second, regulatory clarity. Ethereum spot ETFs were approved in 2024, and staking became the frontier. The Bank of Italy has been clearer about Ethereum's status than about almost any other asset in this market. A European bank cannot move capital without that clarity.
Third, flow alignment. The ETH spot ETF complex has been steadily absorbing inflows. BlackRock clients, per BSCN, rotated roughly $60 million out of IBIT while adding more than $20 million into the ETHA spot Ethereum ETF. The pattern mirrors Intesa's own book: trim Bitcoin, add Ethereum.
The Solana decision is equally informative. Solana staking yields are higher than Ethereum's — often 5 to 8% annual. If Intesa were simply chasing yield, it would have kept the Solana position. Cutting to seven shares while tripling ETH tells you this is a liquidity and settlement decision, not a yield decision. Solana's institutional plumbing — custody, staking infrastructure, regulatory footprint — remains too shallow for a bank of this scale.
You can disagree with that assessment. But the filing is unambiguous about the direction of travel.
The Timing Problem
The timing of Intesa's reduction deserves attention.
The US spot Bitcoin ETF market recorded its worst month on record in June — roughly $4.5 billion in net outflows. That is the exact window captured by this 13F. July reversed: $172.4 million in net inflows, pushing Bitcoin back toward $64,000 in the middle of the month. August has continued the trend with another $170 million so far.
In other words, if Intesa was making a directional bearish call on Bitcoin, it did so at the exact bottom of the institutional flow cycle. The position cut showed up in the June 30 snapshot — the moment of maximum pessimism, right before $4.5 billion of selling turned into fresh inflows.
I have been burned by this kind of timing before. In 2021, during the NFT mania, I staked $15,000 of savings into a high-yield Polygon bridge protocol based on a Discord tip. Lost 60% of principal. Spent three nights reverse-engineering transaction logs on Etherscan. The lesson stuck: sell-side narratives arrive late. The ledger remembers what the code tries to hide.
Apply that lesson here. 13F filings are backward-looking by construction. They arrive 45 days after quarter-end, delayed, netted, and stripped of context. Using them to make a bullish or bearish call on Bitcoin is like trading off last month's weather report.
Contrarian: The Yield Narrative Is the Real Trade
Here is the blind spot in most coverage of this filing: everyone is debating what the IBIT cut means for Bitcoin, and almost no one is asking what the staked ETH accumulation means for Ethereum.
The IBIT story is a distraction. The real signal is the institutional migration toward yield-bearing crypto structured products.
Think about the institutional math. A bank like Intesa cannot hold naked crypto exposure without substantial capital charges. But a staked ETH ETF product, offered by BlackRock, with an audited yield, fits into a framework the bank's risk committee can defend. It produces income. It has a recognized sponsor. It trades on regulated venues.
This is the same dynamic I observed when building volatility arbitrage strategies for institutional desks in Mexico City after the 2024 ETH ETF approval. TradFi risk models are rigid. They misprice crypto-native signals because they lack frameworks for them. But they understand yield. Yield is the bridge.
Bitcoin, from this vantage point, is a directional asset without yield. Staked Ethereum is a yield asset with upside optionality. For a European bank managing risk-weighted assets, the choice is obvious. This is not a comment on which asset will outperform. It is a comment on which asset institutional balance sheets can actually hold.
I also want to address the IBIT puts directly, because the commentary around them has been sloppy. A put position on 500,000 shares, reported alongside a growing staked ETH position, is not evidence of a bank shorting Bitcoin. It is evidence of a bank that wants upside participation without unmanaged drawdown risk. The put premium is the cost of sleeping well under a risk committee's scrutiny.
The options market has been telling us this for months. Institutional open interest in IBIT puts has been climbing as the ETF matured. The marginal buyer of Bitcoin exposure wants protection attached. Intesa's filing is just a public version of a private pattern.
Takeaway: What I'm Watching Next
Three things will confirm or falsify my reading of this filing.
First, the September 13F. If the staked ETH position continues to grow while the IBIT put position is reduced or closed, the rotation thesis is confirmed. If the put position expands alongside more share cuts, the bearish narrative gains credibility.
Second, European peer behavior. Italian banks often lead, but they do not move alone. Watch the next 13F cycle from French, German, and Spanish institutions. If the staked ETH pattern repeats, we are looking at a continental shift in crypto allocation strategy.
Third, the IBIT options market itself. The ratio of put open interest to share volume will tell you whether Intesa's structure is idiosyncratic or systemic.
I trade the gap between expectation and execution. The expectations are set by headlines like "94% IBIT cut." The execution is visible in the options grid, the staking rotation, and the flow reversal that followed the quarter. Those two realities do not agree.
This filing was never about a bank fleeing Bitcoin. It is about a bank learning to price risk in a market that lacks institutional-grade yield products. The ledger remembers what the code tries to hide. In this case, the code is a 13F form, and the ledger is the options grid.
Trust the math, verify the chain, ignore the hype. The hype says Europe is leaving crypto. The data says European institutions are restructuring how they hold it. Those are opposite conclusions, and only one of them survives contact with the filing.