The data shows that 34% of all Bitcoin in circulation has already exposed its public key. That fraction is not a theoretical model. It is a countable property of the current UTXO set. Coins that have moved through P2PK outputs, or that sit in P2PKH change addresses that have been spent once, broadcast their public key the moment the transaction was confirmed. From that point forward, the private key is the only barrier between the funds and anyone with enough computational leverage. No quantum computer is required to list the exposed coins. A block explorer suffices. The second half of the attack—deriving a private key from an exposed public key—requires Shor's algorithm at a scale that the public research record says does not exist. But the first half is irreversible. This week's CNBC interview did not introduce a new vulnerability. It introduced a latency test. Jim Cramer told millions of viewers he was selling Bitcoin. No wallet address, no amount, no signed message. The network's actual exposure was already waiting on-chain.
Context is necessary. IBM and the University of Chicago recently published a 70-logical-qubit execution using 468 T gates and a runtime of roughly 16 minutes. The word logical matters. Physical qubits are noisy; logical qubits are error-corrected units that actually preserve information. A 70-logical-qubit machine is a meaningful hardware marker. It is not a codebreaker. The gap between this result and a secp256k1 compromise is not incremental. Google Quantum AI, Stanford, and the Ethereum Foundation have jointly estimated that breaking secp256k1 would require 1,200 to 1,450 logical qubits and 70 million to 90 million Toffoli gates. The difference is roughly 20x on qubits and five orders of magnitude on gates. That is a chasm. But a chasm is also a corridor: its width can shrink.
Inside that corridor, three things happened at once. First, IBM CEO Arvind Krishna said in the CNBC exchange that quantum computing would produce real commercial impact within three to four years, and linked the timeline to his company's revenue projections. Second, Jim Cramer, whose public Bitcoin record includes dismissing the asset at $16,796 in December 2022, announced that he was selling his Bitcoin. He did not say when, how much, or to which address. Third, BIP-361 sits in draft. The proposal, authored by Jameson Lopp of Casa and five others, is the first serious attempt to make wallet software recognize public key exposure. It carries a baseline figure that should stop every complacent conversation: as of March 1, 2026, more than 34% of Bitcoin has already leaked its public key.
Bull market psychology amplifies noise and filters signal. A CNBC warning about quantum breaking Bitcoin is noise that trades on an older fear. The actual signal is the 34% figure and the lack of a governance mechanism to move it before the hardware matures. The rest is narrative. This article is not about the narrative. It is about the ledger.
The Gap Is a Function, Not a Date
Let me start with the arithmetic, because arithmetic is the only thing here that does not lie. IBM's experiment executed a circuit with 70 logical qubits and 468 T gates. Google's estimate for factoring a secp256k1 public key calls for 1,200 to 1,450 logical qubits. Equivalent logical qubits are not directly comparable to T gates, but the engineering distance is still brutal. The gate count sits at 70 million to 90 million Toffoli gates, or roughly 70 million to 90 million of the operations that quantum factorization requires. Five orders of magnitude separate a 468-gate experiment from a 70-million-gate computation. A reviewer who understands fault tolerance looks at that number and reads it as a timeline, not an impossibility. The trajectory matters more than the current absence.
In 2018, I spent three months auditing the 0x Protocol v2 smart contracts. I found seven critical issues in the order routing logic. One of those issues was not reachable with the compiler and tooling of that exact moment. It was a deferred vulnerability, not a false positive. I flagged it because exploitability is a function of time, not of current tooling. Quantum computing has the same shape. The 70-logical-qubit experiment does not break Bitcoin. It breaks a previous psychological baseline. Anyone who claims the threat is solved because IBM is not Google is reading the wrong stage of the curve.
The IBM result is a statistical lower bound on hardware fidelity. It proves that error-corrected circuits can run for a meaningful duration and produce an output that survives decoherence. It does not prove that the same hardware can run Shor's algorithm at curve scale. The two tasks share the word quantum, but the engineering requirements differ by orders of magnitude. Gate errors need to fall below a threshold that current systems still routinely miss. The logical-qubit count needs to expand by more than an order of magnitude. The latency budget needs to handle millions of gates before noise collapses the computation. Any one of those requirements alone can take years. Together, they are a multi-year experiment.
Code speaks louder than promises. Bitcoin's code still uses secp256k1. The whole network's security rests on a 128-bit elliptic curve that everyone can name. That design was reasonable when the cheapest attack was a brute force over 2^128 keys. It remains reasonable today because no public machine can run the required factorization. But the margin shrinks with every logical-qubit milestone. The correct posture is not panic. It is migration planning. The earlier the community starts, the less likely the timeline is to become a cliff.
The CEO's Calendar Has a Ticker
Arvind Krishna is not an independent auditor. He is the chief executive of a company that sells quantum computing services and hardware. In the same interview cycle, he tied IBM's earnings growth to quantum commercialization before 2028 or 2029. That statement is a financial forecast. It is also a marketing timeframe. When a CEO says a technology will be useful in three to four years, and his revenue model depends on that technology reaching utility before the next earnings horizon, the number is not a scientific constant. It is a target. I have seen this pattern in nearly every protocol I have reviewed. The person selling the shovel has the most optimistic surveyor's report. It is not fraud. It is perspective. The community should weight estimates based on incentive, and the incentive skew is visible.
Google, Stanford, and the Ethereum Foundation have no comparable commercial position in quantum hardware. Their joint estimate is still uncertain, but it is structurally more neutral. They put the requirement at 1,200 to 1,450 logical qubits. The distance from IBM's 70 logical qubits is roughly 17x to 20x, and the gate count gap is five orders of magnitude. I do not cite the gap to dismiss the threat. I cite it to correct the pitch. The market heard three to four years. The research literature says not yet, and not with the current public roadmap. Those are different inputs for any risk model.
Follow the gas, not the narrative. On-chain activity, not television statements, is the measure of conviction. If Cramer's announcement were a real liquidation, the withdrawal patterns at major exchanges would show it. There were no reported large transfers, no custody outflow, no on-chain signature proving he held the wallet. A sell declaration without a transaction hash is a media event. The price action of Bitcoin after the interview did not show a liquidation cascade. It showed a market that has learned to classify Cramer as a signal jammer rather than a signal. The only unquantified movement is in the noise layer of retail chat rooms.
BIP-361 Is the Real Ledger
The BIP-361 draft is the only piece of the current discussion that is both technical and actionable. Jameson Lopp and five co-authors put it forward. It is not active. It has not been merged into Bitcoin Core. It is a draft standard that would make wallet software aware of public key exposure. That seems like a modest goal. It is actually a precision instrument for an uncomfortable truth.
Let me be explicit about what the 34% baseline means. Every Bitcoin user who holds coins in a legacy address that has moved once has already shown their public key. P2PK addresses broadcast their public keys at rest. P2PKH change addresses leak their public key whenever the change output is spent. These are not exotic corner cases. The spent-change pattern is the default behavior for a large fraction of older wallets. The fraction, measured by supply, is 34% as of March 1, 2026. That includes some of the most dormant coins in the network. The coins that have not moved in years are overrepresented because they were created in the era of P2PK and P2PKH wallets. The very addresses most likely to be labeled cold storage are the least safe in a quantum epoch.
Here is the blind spot that my forensic work keeps finding. Dormant coins are not a static liability; they are an unmanaged liability. Once a public key is exposed, the coin cannot be un-exposed. The only remedy is to move the coin to an address whose public key remains hidden. That means spending the UTXO and replacing it with a P2TR or other taproot output. Friction, however, is real. Users need to understand the reason. Wallets need to support the migration without creating address-reuse mistakes. Exchanges need to accept the new output types and update their internal accounting. Every day of delay adds to the population of exposed coins. The number does not shrink on its own.
Trust is verified, not given. The Bitcoin protocol does not authenticate identity. It authenticates signatures. Once a public key is on-chain, the only trust left in the system is the computational difficulty of inversion. That trust decays as quantum processing improves. A migration standard like BIP-361 is an early warning system. It does not solve the math. It tells the operator which doors are locked and which are only closed. That is why the 34% number is the most useful metric in this entire conversation. It converts a vague fear into an address-level inventory.
Rules Without a Ruler
Regulatory timelines now intersect with a protocol that has no legal personality. NIST's draft guideline proposes to ban 128-bit curves after 2035. secp256k1 is a 128-bit security curve. The Hong Kong Monetary Authority has instructed banks to reach quantum readiness by 2030. These are not recommendations for crypto users. They are binding constraints for licensed custodians, banks, and ETF trustees. The consequence is indirect but measurable. A licensed bank that holds Bitcoin will need to answer its auditor's question: what is your quantum exposure? The honest answer is that Bitcoin's entire current curve is at 128-bit security, and 34% of the supply has already exposed its public key. A regulator may not demand an immediate solution. An auditor will demand a documented plan.
The deeper issue is governance. Bitcoin has no CEO, no board, and no legal entity that can sign a remediation plan. It is, in effect, a DAO without a jurisdiction. Most DAOs have no legal status, and when a standard must be maintained, the absent counterparty becomes the risk. NIST can set a 2035 deadline for federal systems. HKMA can set a 2030 deadline for banks. Neither can force Bitcoin's miners, core developers, and wallet maintainers to adopt a new signature scheme. The network can only move by rough consensus and soft fork. That process takes years. Every soft fork in Bitcoin's history required a coordination period measured in quarters, not months. A post-quantum migration requires new address formats, new key generation algorithms, new signing schemes, and exchange-side upgrades. The conservative estimate is five to ten years of interoperability.
Regulators are not ignorant of the technology. They have an incentive to keep the rules vague. A precise rule would force them to define a migration deadline for a protocol that has no accountable actor. The NIST timeline is precise because it applies to federal procurement. The SEC has no equivalent guidance because it prefers to regulate through enforcement. The ambiguity is a feature, not a bug. In practice, the pressure will land on the weakest link: the custody chain. That is where the engineering has to start.
Cramer Is Noise With a Timestamp
The Cramer episode deserves a paragraph because it reveals how the market absorbs a technical story. Cramer said he was selling. He did not confirm the sale, did not disclose the amount, and did not provide an address. The traditional financial system treats an aggregation of unverified statements as sentiment. On-chain analysis treats them as absent data. I follow the chain. The chain shows no evidence of a meaningful liquidation tied to the interview. The price action was consistent with shallow volatility, not supply shock.
The inverse Cramer trade is also statistically dead. Tuttle's Inverse Cramer ETF produced a -15.7% return during its comparative window while SPY delivered +25.4%. A simple strategy of always trading against Cramer lost. The 2012 Management Science paper found something narrower: a 2.4% overnight bounce after a bullish recommendation, followed by a complete reversal within twelve trading days. That is not a directional signal. It is a measure of retail latency. The trade is to short the overnight enthusiasm, not to place a long-term bet against a television personality. In a bull market, even that trade can be crowded.
I have watched this dynamic since the 2021 NFT wash trading investigation. The lesson is the same. When a public figure makes a statement without a transaction hash, the correct response is a liquidity audit, not a debate about their credibility. The market is not responding to Cramer. It is responding to the gap between his words and the ledger. That gap is where sophisticated participants operate.
What the Bulls Got Right
None of the above means the bulls are wrong to hold Bitcoin. The baseline probability of a quantum break within three years is low. The existing ecosystem liquidity, regulatory acceptance, and historical security of Bitcoin make it irrational to rotate into an untested quantum-safe chain. In my 2024 ETF custody review, I saw how slow institutional adoption is when the standard is mere compliance. A chain that has never been tested under adversarial conditions is a much larger audit finding than a curve that has endured a decade. The cost of switching wins is not justified by the current margin of quantum improvement.
But the contrarian angle has a second layer. The margin of improvement is not linear. Google's own estimate for the qubit requirement has already improved by roughly a factor of 20 from earlier figures. What was once a 25,000-logical-qubit problem is now a 1,200-logical-qubit problem. That is not a step function in the threat; it is a compression of the timeline. The market's current indifference is rational on a years-to-next-hardware-milestone basis, but it ignores the fragility of the exposed 34%. Those coins will not wait for consensus. They will become a reserve that nobody can spend. The longer migration is delayed, the more the digital gold narrative depends on an assumption that has a visible expiry date.
There is also a self-healing path. If Bitcoin adopts a post-quantum signature scheme through a careful soft fork, the network would be more durable than most traditional banks. The event could become a bullish narrative: the oldest blockchain demonstrating that it can upgrade its own cryptographic foundation. But that requires the same community that could not raise the block size to align on a new curve. I would not price that outcome into today's portfolio.
The takeaway is operational, not directional. If you hold Bitcoin in an address that has been moved even once, you are part of the 34%. You can act before the deadline. Move your funds to a P2TR address. Verify that your wallet supports it. Then stop treating quantum FUD as news and start treating it as a checklist. The protocol has no legal entity to confirm compliance. You are the compliance layer for your own coins. Logic outlives the hype cycle. Code speaks louder than promises. The question is not whether quantum computing arrives. The question is whether your wallet will be on the safe side of that ledger.