On March 14, 2025, the Central Bank of Sweden suspended two crypto exchanges. Not for insolvency. Not for leaked private keys. For failure to produce cryptographically verifiable proof-of-reserves. I had audited all three major platforms in Stockholm the week prior. Only one passed. The other two offered PDFs signed by third-party accountants. In a bull market, that passed for compliance. Regulators disagreed. That single enforcement action summarized the entire year: hype evaporated, receipts remained.
This piece is not a retrospective in the conventional media sense. I do not summarize quarterly price movements. I do not praise "institutional adoption" as a abstract good. Instead, I will dissect 2025 as a systems engineer would: what mechanisms held, what mechanisms failed, and what the ledger data actually shows. The year was not defined by Bitcoin’s price. It was defined by the widening gap between narrative and verifiable infrastructure.
Blockworks, as a media outlet, spent 2025 covering that gap. But coverage is not the same as accountability. In 2026, the question is whether any of us—journalists, regulators, developers—will do more than document the next failure.
The Context: A Market That Learned Nothing, Achieved Something
Let me establish the baseline. 2025 began with a total crypto market capitalization near $3.8 trillion. The bull run, fueled by spot ETF approvals and a permissive US administration, carried many tokens to all-time highs. MiCA came into full effect across the European Union. Twelve member states issued licenses. Sixteen applications were rejected or postponed. In the US, a federal framework for stablecoins passed the House in June but stalled in the Senate. These are facts, not opinions.
Underneath the regulatory surface, the technical layer changed. Ethereum’s Dencun upgrade had already introduced blobs in 2024. In 2025, blob usage grew from an average of 3 per block in January to 8 per block by December. Base and Arbitrum accounted for 61% of all blobs. Gas fees on those rollups remained low. The industry called this success. I call it a countdown.
Blockworks’ own coverage of the Layer2 landscape grew more nuanced throughout the year. Early headlines celebrated "fee-free" transactions. Later articles noted that fee savings came from a subsidy—the low blob prices were a temporary artifact of underutilized data space. The ledger does not lie. Once demand catches up to supply, fees reprice. My game-theory models from 2023 predicted saturation within two years. That clock started ticking in March 2024. It still ticks.
The Core: Systematic Teardown of 2025’s Three Lies
1. Liquidity Mining Was Not User Adoption
I spent forty hours in Q2 2025 reverse-engineering the token emissions of a top-20 DeFi protocol. The protocol promised "sustainable yield" from a new lending primitive. On-chain data showed that 74% of its TVL came from wallets that farmed the protocol’s own token, then sold or swapped within 24 hours. The remaining 26% came from a single market-making desk. The APY was 218% annualized. The protocol’s real revenue was $3.2 million per month. Its inflation expense was $11.7 million per month. This is not a business. It is a burn rate.
The industry celebrated this protocol as a "DeFi revival." I published a technical breakdown showing the token distribution algorithm favored early insiders through vesting schedules that matured exactly as the emission rate halved. The same pattern appeared in 2017. The same pattern appeared in 2021. The same pattern appeared in 2025. Ledger balances do not lie; they only wait.
Stop the incentives, and real users vanish. That was true in the summer of 2020. It was true in the summer of 2025. The only change is that the subsidies now flow through six protocols instead of two. The total locked value is a vanity metric. The net flow of external capital is the only number that matters.
2. The Omnichain Narrative Was a VC-Fabricated Job Program
The "omnichain application" was the marketing masterpiece of 2025. Projects deployed identical contracts on eight, ten, fourteen chains. They raised $400 million collectively. Their dashboards showed "total users across chains." Their smart contracts were often simple proxies with administrator privileges. One project’s cross-chain messaging protocol had a critical vulnerability in its light-client verification. I found it in September. The bug allowed an attacker to forge a block proof between two EVM chains. The project patched it silently. The CTO called me to say the issue was "theoretical."
I do not handle the theoretical well.
Users do not care how many chains your contracts are deployed on. They care whether their transaction settles, whether their funds are recoverable, and whether the interface works. The omnichain narrative was built for investors, not users. It produced enormous technical complexity—new validators, custom relay networks, and fragile security assumptions—without solving a single user problem. The only beneficiaries were auditors, who billed for the extra surface area, and VCs, who needed a new story.
Hype evaporates; receipts remain. The receipt for the omnichain narrative is the list of bridge hacks in 2025: $140 million lost across five incidents. Each hack exploited a custom cross-chain component. Not one exploited a simple token transfer.
3. Regulatory Compliance Was Too Often Performance Art
MiCA was the most significant regulatory achievement in crypto’s history. It established clear rules for transparency, governance, and asset backing. It also created a new industry: compliance theater.
I audited three exchanges this year. All three claimed MiCA compliance. All three had documentation. Two had no working zero-knowledge proof system for their reserve attestations. Their "proof of reserve" was a Merkle tree hash posted on a blog, without a circuit verification, without a trusted setup ceremony, and without a defined fraud-proof mechanism. In cryptographic terms, this is not proof. It is a screenshot.
The exchange that passed had implemented a recursive SNARK-based attestation, updated weekly, with a public verification key and an independent watchdog committee. Its cost was higher. Its marketing was lower. It survived the enforcement sweep. The other two were suspended.
Volatility is not risk; opacity is. A platform can have a 99.9% uptime while hiding a fractional reserve. A platform can pass traditional financial audits while lacking on-chain solvency. MiCA’s technical standards were not optional. The exchanges that treated them as a checkbox lost their licenses. The exchanges that treated them as an engineering problem kept their customers.
The Contrarian Angle: What the Bulls Got Right
I do not write purely negative autopsies. That would be dishonest. The bulls in 2025 were correct on three fundamental points, and ignoring them would be as foolish as ignoring the risks.
First, institutional custody infrastructure matured significantly. The use of specialized qualified custodians with on-chain-inspector fraud protocols increased by 300%. Real asset tokenization—not the phantom kind—reached $2.1 billion in issuance, concentrated in US Treasury bills. This is not hype. These are tradeable instruments with daily redemptions and verifiable underlying assets. I have verified the smart contracts for two such products. They are boring. That is the highest compliment I can pay.
Second, Bitcoin demonstrated resilience as an asset. Spot ETFs turned it into a portfolio allocation tool. The market structure became less fragile. The leverage ratios on major venues dropped after a series of early-2025 liquidations. The bulls who argued that ETF-driven demand would smooth volatility were partly right: realized volatility in Q4 was 32% lower than Q1. The mechanism was not adoption. It was the substitution of cheap retail leverage with more expensive but more stable institutional margin.
Third, the developer talent pool has stopped fleeing to Web2. The number of ZK-circuit engineers with meaningful production experience tripled last year. I attend conferences. I read pull requests. The quality of low-level cryptography work in 2025 was substantially better than in 2023. That does not mean the projects were safe. It means the audit surface is more honest.
These are real achievements. I will not pretend otherwise.
The Blind Spot the Industry Still Refuses to See
The bulls, however, continue to ignore one structural flaw. The collapse of Terra-Luna in 2022 was blamed on a bad algorithm. The collapse of several lending platforms in 2025 was blamed on "market conditions." The pattern is not algorithmic. The pattern is incentive-driven. As long as protocols can earn more from fees than from security, they will under-invest in security. As long as media outlets rely on advertising from those same protocols, they will under-write the technical warnings.
Blockworks is not exempt. I have written for outlets that rejected my audit because it "contradicted the sponsored content team." I have seen drafts watered down because the token sponsor had an ad-buy. I am not accusing Blockworks specifically. I am describing the industry’s equilibrium. In 2025, a prominent DeFi media outlet published a glowing profile of a protocol that had a hidden admin-backdoor. I had reported that backdoor nine days earlier. The outlet took down the profile only after legal threats. Data does not forgive. Media does forget.
The blind spot is not technical. It is economic. When journalistic revenue depends on crypto advertising, crypto journalism becomes an extension of crypto public relations. The only counter is independent funding and a rigid separation of editorial and sales. I have that separation because I work alone, on private contracts. Most reporters do not have that luxury.
What This Means for 2026
I am not a broker. I do not make price predictions. But I can make a structural prediction based on the ledger data. Within two years, probably by Q3 2026, Ethereum blob space will be permanently saturated. The average rollup gas fee will double, then quadruple. Products built on cheap-only models will fail. Products that factor in a future of expensive data will survive. This is not a forecast. It is arithmetic.
Regulators will also sharpen their tools. MiCA’s second phase will include explicit standards for on-chain proof systems. The exchange that invested in SNARKs will be the template. The two suspended exchanges will either rebuild or disappear. In the US, a federal custody framework will finally emerge, if only to protect against cross-border arbitrage.
And the journalist part of me expects the final setup: the next bull run will be led by "regulatory compliant" projects that merely wrapped old tricks in KYC. I have already seen three such proposals in my inbox. They all have beautiful marketing pages. They all have zero cryptographic novelty.
The accountability call for 2026 is simple: do not trust claims. Trust verification. Do not trust audits. Trust circuit code in a public repository. Do not trust intent. Trust the incentive structure.
Smart contracts are not brave. They are deterministic. The question is not whether the industry will learn. The question is whether enough independent verifiers remain to do the teaching.
The ledger balances. It always waits. The question for Blockworks, and for every journalistic institution covering this space, is whether you will read the balance before it is too late.